Inspect the sources.
Search every product claim and open the first-party page that supports it.
35 active products779 primary sources8 watchlist records0 affiliate sources
37 of 37 products, all statuses
Claude CodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Git worktree, Managed sandbox
Persistent memory
Proprietary
Claude Code overviewSupported surfaces, CI integrations, MCP, cloud sessions, and remote workflowsdocsHow Claude Code worksAgentic loop, built-in tools, context management, and host interaction modeldocsPlatforms and integrationsCLI, Desktop, IDE, web, mobile, Slack, and CI surface comparisondocsProduct pageOfficial product surfaces, install paths, Pro and Max plan inclusion, and feature announcementsannouncementCLI referenceTerminal, print mode, permissions, MCP, and automation flagsdocsRun programmaticallyNon-interactive print mode, Agent SDK entrypoints, and headless automation constraintsdocsTools referenceBuilt-in tool inventory, per-tool permission requirements, and execution behaviordocsSecurityRead-only defaults, sandbox Bash, working-directory boundary, prompt-injection and cloud safeguardsdocsSandboxingOpt-in OS sandbox, fail-open default, scope, escape hatch, platforms, and limitationsdocsSandbox environmentsBash sandbox vs runtime vs devcontainer vs VM vs web isolation boundaries and enforcementdocsDevelopment containersContainer isolation, egress firewall example, managed policy, and unattended skip-permissions usedocsPermissionsTool rules, approval modes, managed restrictions, sandbox interaction, and bypass controlsdocsPermission modesManual, acceptEdits, plan, auto, dontAsk, and bypassPermissions tradeoffs and protected pathsdocsConfigure auto modeClassifier trusted infrastructure, hard deny rules, environment slots, and org overridesdocsAuto mode announcementClassifier-backed autonomy as a safer alternative to dangerously-skip-permissionsannouncementAuto mode engineering evaluationClassifier design, 10,000-session production evaluation, real and synthetic false-negative rates, false positives, and high-stakes limitationannouncementParallel agentsSubagents, agent view, experimental teams, batch delegation, and worktree choicesdocsAgent viewResearch-preview background session manager, parallel dispatch, worktrees, permission inheritance, local lifecycle, and quota scalingdocsSubagentsSpecialized delegated contexts, per-agent tools, permissions, hooks, and worktreesdocsAgent teamsExperimental coordinated sessions, shared tasks, messaging, permissions, and known overheaddocsDynamic workflowsScripted multi-subagent orchestration for audits, migrations, and cross-checked researchdocsWorktree isolationParallel session and subagent file isolation, base refs, cleanup, and boundariesdocsCheckpointing and rewindAutomatic file checkpoints, rewind, conversation restore, and external-side-effect limitationsdocsSession managementLocal session persistence, resume, naming, branching, export, and retention controlsdocsProject memoryCLAUDE.md hierarchy, default-on local auto memory, storage, loading limits, and user auditdocsHooks referenceLifecycle hooks, pre-tool enforcement, subagent events, transcripts, and managed controlsdocsMCP integrationLocal and remote MCP servers, authentication, scopes, resources, and security warningsdocsManaged MCP accessOrganization allowlists, denylists, and managed MCP server restrictionsdocsPlugin marketplacesPlugin distribution, agents, hooks, MCP and LSP components, source pinning, updates, validation, and managed marketplace restrictionsdocsChrome integrationBeta visible-browser automation, site permissions, authenticated state, and plan restrictionsdocsComputer useCLI computer-use for native apps, screen control, and host GUI automation boundariesdocsClaude Code on the webDurable cloud sessions, isolated VMs, network policy, credential proxying, handoff, and limitsdocsDesktop applicationDesktop sessions, parallel worktrees, previews, computer use, and enterprise configurationdocsVS Code extensionIDE integration, inline diffs, permission modes, and extension security and privacydocsJetBrains IDEsJetBrains plugin surfaces, selection context, and terminal-backed Claude Code sessionsdocsRoutinesScheduled local and remote tasks, credentials, result delivery, and operational constraintsdocsGitHub ActionsCI automation, pull-request and issue triggers, authentication, permissions, and securitydocsManaged Code ReviewResearch-preview multi-agent pull-request review, verification and deduplication, triggers, plan scope, neutral checks, and optional user-defined CI gatingdocsGitLab CI/CDGitLab pipeline automation, authentication, and CI security controls for Claude CodedocsSettingsUser and project settings hierarchy, sandbox keys, permission defaults, and managed deliverydocsServer-managed settingsCentral policy delivery, precedence, fail-closed refresh option, audit hooks, and limitationsdocsAdmin setupEnterprise rollout map for providers, managed policy, monitoring, and data handlingdocsAuthenticationClaude subscription, Console API, enterprise providers, credentials, and long-lived tokensdocsFeature availabilityFeature matrix across subscription plans, Console, Bedrock, Agent Platform, and FoundrydocsEnterprise network configurationProxy, custom CA, mTLS, and required network domains for enterprise deploymentsdocsLarge codebasesMonorepo and large-tree guidance with nested CLAUDE.md, sparse worktrees, and package focusdocsUsage monitoringOpenTelemetry metrics and events, configuration, privacy boundaries, and dashboardsdocsData usageWhat Claude Code transmits, retention posture, telemetry services, and opt-out controlsdocsZero Data RetentionQualified Enterprise enablement, direct-inference scope, excluded product surfaces and integrations, and model availability changesdocsAgent SDK overviewProgrammatic agent loop, tools, permissions, sessions, subagents, MCP, and deployment surfacedocsAgent SDK secure deploymentIsolation, credential management, and network controls for hosted Agent SDK agentsdocsRuntime errors and retriesRuntime failure classes, automatic retries, unattended watchdog behavior, timeouts, partial responses, recovery, and provider-specific errorsdocsClaude Code changelogVersioned release notes for features, permission modes, sandboxing, and surface changesdocsClaude Code 2.1.220 releaseCurrent public distribution release used for product-version verificationannouncementClaude Code public repositoryRelease artifacts, plugins, examples, support automation, license, and absent core sourcerepositoryAnthropic sandbox runtimeOpen Seatbelt and bubblewrap runtime for process filesystem and network isolationrepository
CodexCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Sandbox-first
OS sandbox, Git worktree, Managed sandbox
Persistent memory
Apache-2.0 client
Codex CLIInteractive terminal client, installation, authentication, sessions, and supported workflowsdocsDeveloper command referenceCLI, exec mode, approvals, sandbox flags, and local providersdocsApprovals and securityDefault local sandbox, approval policies, cloud isolation, network controls, and threat boundariesdocsSandboxingNative command isolation across desktop, CLI, and IDE, filesystem and network boundaries, and approval escalationdocsAuto-reviewReviewer-agent handling of eligible sandbox escalations, unchanged boundaries, circuit breakers, overrides, and non-deterministic limitsdocsPermission profilesBeta least-privilege filesystem and network profiles plus legacy-mode incompatibilitydocsPermission modesApproval-oriented, automatic, full-access, and custom local action modes across desktop, CLI, and IDE surfacesdocsCommand rulesExperimental prefix rules for allow, prompt, or forbid decisions when commands request execution outside the sandboxdocsWindows sandboxNative Windows filesystem and network isolation, PowerShell operation, setup, protected paths, and troubleshootingdocsSubagentsDefault parallel delegation, custom agent roles, tools, models, and token overheaddocsCodex memoriesLocal persistent memory store, controls, scopes, and distinction from required project rulesdocsAGENTS.md instructionsGlobal and project instruction discovery, nested overrides, precedence, size limits, and code-review guidancedocsExecution environmentsLocal, Git-worktree, and cloud execution modes in the desktop appdocsGit worktreesParallel checkout isolation, scheduled-task worktrees, handoff, and Git-only constraintsdocsLocal environmentsDesktop worktree setup scripts, shared project actions, generated configuration, and execution timingdocsCodex cloudDelegated cloud tasks, isolated environments, repository workflows, and review surfacedocsCloud environmentsContainer creation, dependency setup, default-off agent internet, AGENTS.md use, validation loop, and diff handoffdocsCloud internet accessDefault offline agent phase, domain allowlists, methods, secrets, and exfiltration risksdocsBrowserShared browser surface, separate profile, site interaction, downloads, and safety guidancedocsComputer UseDesktop GUI control, plugin setup, OS permissions, attended review, and external side effectsdocsRemote connectionsMobile and desktop steering of connected hosts, inherited credentials and permissions, SSH projects, approvals, and reviewdocsNon-interactive modecodex exec scripting, CI operation, structured output, sandbox, and approval settingsdocsLifecycle hooksPre and post tool gates, validation, logging, subagent events, trust, and side-effect limitsdocsModel Context ProtocolLocal and remote MCP server supportdocsSkills and pluginsReusable instruction packages, installable bundles, connector-backed MCP tools, resources, and supported surfacesdocsPluginsPlugin discovery, installation, supported ChatGPT and Codex surfaces, connected tools, and availability controlsdocsIDE extensionVS Code-compatible, Xcode, and JetBrains integrationsdocsChatGPT desktop appDesktop project, file, long-running work, parallel chat, worktree, browser, plugin, and scheduled-task surfacedocsAuthenticationChatGPT, API-key, and enterprise access-token authentication, surface availability, credential storage, and managed restrictionsdocsCodex GitHub ActionCI installation, codex exec, patch and review workflows, secrets, and safety strategydocsCode reviewRead-only review of local or branch changes, prioritized findings, review pane, and user-controlled staging or revertdocsGitHub code reviewPull-request review setup, manual and automatic triggers, AGENTS.md rules, permissions, and posted GitHub findingsdocsCodex SecurityOptional security plugin and cloud workflows for repository scans, code-change review, finding triage, fixes, and reportsdocsCodex SDKProgrammatic threads, structured events, CI integration, and server-side constraintsdocsCodex App ServerEmbeddable protocol, persistent JSONL threads, streaming events, and deprecated context rollbackdocsScheduled tasksRecurring background tasks, local projects, dedicated worktrees, status, and run historydocsManaged configurationAdmin-enforced requirements for permissions, sandbox, hooks, MCP, plugins, and precedencedocsAdmin rollout guideEnterprise rollout boundaries for workspace access, local runtime policy, cloud, APIs, plugins, connectors, and source systemsdocsAdvanced configurationCustom and local providers, OpenTelemetry, profiles, network proxy, and configuration scopedocsAmazon Bedrock providerOpenAI models through Bedrock, AWS-native authentication, compatible local surfaces, request path, and policy boundariesdocsFeature maturityOfficial under-development, experimental, beta, and stable labels with change and production-use expectationsdocsWhat's newWeekly dated digest of ChatGPT and Codex surface changes with links to current product documentationdocsOpen-source component mapPublic CLI, SDK, app server, universal cloud image, and proprietary surface boundariesdocsCodex CLI 0.145.0 releaseCurrent stable CLI distribution release used for product-version verificationannouncementCodex repositoryPinned 0.145.0 CLI, SDK, app-server, sandbox, tests, workflows, and Apache-2.0 licenserepositoryRollout trace implementationPinned implementation of local rollout inspection used as code-verifiable trace evidence rather than a benchmark resultrepository
OpenCodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
No first-party isolation
Session-based
MIT
OpenCode introductionTerminal, desktop, IDE, installation, and basic provider setupdocsProvidersMore than 75 cloud, subscription, gateway, OpenAI-compatible, and local providersdocsModelsProvider and model selection, local models, variants, defaults, and model-quality warningdocsAgents and permissionsPrimary agents, parallel subagents, built-in roles, compaction, models, and per-agent permissionsdocsPermission policyAllow, ask, deny, permissive defaults, command patterns, external paths, and auto modedocsTerminal interfaceSessions, compaction, export, sharing, and Git-dependent undo and redodocsCommand-line interfaceInteractive, run, server, attach, session fork, auto approval, and automation commandsdocsBuilt-in and custom toolsHost tools, web retrieval, permission controls, custom tools, MCP, and ignore behaviordocsMCP serversLocal and remote MCP transports, OAuth, enablement, environment, permissions, and context costdocsPluginsLocal and npm plugins, hooks, custom authentication, tool interception, and execution scopedocsSkillsProject and user skills, discovery, permissions, metadata, and lazy loadingdocsRules and project instructionsAGENTS.md hierarchy, reusable instructions, compatibility files, and context scopedocsLSP integrationLanguage-server diagnostics, supported servers, custom configuration, and disablementdocsHeadless serverOpenAPI server, localhost default, optional Basic auth, event stream, and client APIsdocsLocal web interfaceBrowser client, local binding, optional password, network exposure, CORS, and shared statedocsSDKTyped programmatic client, server lifecycle, sessions, messages, events, and automationdocsGitHub automationIssue and pull-request agents, schedules, workflow permissions, secrets, branches, and reviewsdocsConversation sharingManual, automatic, and disabled sharing, public-link retention, deletion, and privacy guidancedocsEnterprise deploymentCentral configuration, SSO, gateway restrictions, data path, sharing caveat, and roadmap limitsdocsProvider policiesExperimental provider allow and deny policy, precedence, wildcard matching, and scopedocsGit snapshot implementationSnapshot storage, Git-only enablement, ignored and large-file exclusions, restore, and pruningrepositorySession revert implementationMessage undo and redo, snapshot restoration, patch reversal, and session-state cleanuprepositoryOpenCode 1.18.5 releaseCurrent stable distribution release used for product-version verificationannouncementOpenCode source treePinned 1.18.5 implementation, security policy, tests, workflows, clients, and licenserepositoryPinned security policyExplicit absence of a sandbox, permission-policy boundary, local server authentication default, and security disclosure processrepository
Pi AgentCoding harness; Extensible harness
Coding harness
Extensible harness
Single-agent loop
Host-first
No first-party isolation
Session-based
MIT
Pi Agent Harness repositoryPinned 0.82.1 source, minimal core, package layout, test assets, license, and security posturerepositoryPi 0.82.1 releaseCurrent stable package release and distribution version verificationannouncementQuickstartInstallation, terminal workflow, provider connection, built-in tools, and basic session usedocsCommand-line usageInteractive, print, JSON, RPC, session, model, tool, and resource-loading flagsdocsProvidersSubscription logins, API providers, custom endpoints, and local-model runtimesdocsLocal llama.cpp modelsBundled local-model router, model download and lifecycle, context, and device configurationdocsCustom providersOpenAI-compatible and custom endpoint configuration, authentication, headers, and model metadatadocsSecurityHost execution, absence of permission prompts and sandbox, project trust, and external isolation patternsdocsContainerizationWhole-process Docker isolation pattern and distinction from built-in harness controlsdocsSessionsAutomatic JSONL persistence, resume, branching, forks, cloning, export, and sharingdocsSession formatTree-structured JSONL entries, parent linkage, tool results, metadata, and versioningdocsCompactionAutomatic and manual context compaction, summaries, branch handling, and information lossdocsJSON event streamMachine-readable headless events for integrations and custom interfacesdocsRPC modeBidirectional JSONL control, prompts, tools, session statistics, and host integrationdocsSDKEmbeddable agent sessions, model runtime, session management, tools, and event subscriptionsdocsExtensionsTypeScript hooks, tools, commands, UI, custom workflows, trust, and code-execution scopedocsSkillsProject and user skills, discovery, validation, and instruction loadingdocsPackage ecosystemInstallable extension bundles, local packages, distribution, trust, and update behaviordocsPi evaluation packageProject-owned behavioral evaluation harness, provider selection, and non-independent scoperepository
Oh My PiCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
Git worktree
Persistent memory
MIT
Oh My Pi repositoryPinned 17.1.7 source, LSP, debugger, browser, subagents, tests, and MIT licenserepositoryOh My Pi 17.1.7 releaseCurrent stable package release, tool-input rewriting, schema revalidation, approval routing, and distribution version verificationannouncementProduct overviewInstallation, terminal entry point, provider breadth, tools, agents, and extension surfacedocsApproval modesDefault yolo posture, tiers, per-tool policy, destructive overrides, ACP, and subagent behaviordocsSettings referenceApproval defaults, feature gates, project settings, memory, MCP, and task isolationdocsProvider referenceCloud subscriptions, API providers, local runtimes, model roles, and fallback configurationdocsRPC protocol referenceNDJSON automation, host tools, UI requests, and subagent event streamsdocsTask subagentsParallel task batches, typed results, background execution, agent discovery, and workspace isolationdocsBrowser toolPuppeteer and CDP control, Chromium download, applications, screenshots, network, and process effectsdocsMCP configurationLocal and remote transports, OAuth, profiles, discovery, credentials, timeouts, and enablementdocsCheckpoint toolDisabled-by-default conversation marker, in-memory scope, persistence, and absent file snapshotdocsRewind toolConversation pruning, retained report, branch persistence, and absent filesystem recoverydocsAutonomous memoryOptional persistent project memory, extraction, consolidation, redaction, scope, and defaultsdocsACP implementationAgent Client Protocol entry point, editor integration, runtime options, and headless surfacerepositoryEvaluation runtimeProject-owned execution kernels and engineering tests rather than independent coding outcomesrepositoryRuntime hooksCurrent extension-backed hook path, lifecycle events, pre-tool blocking or input rewriting, fail-closed interception, and loading behaviordocsRuntime extensionsIn-process extension runtime, tools, handlers, tool-input rewriting with schema and approval rechecks, session APIs, timers, and absent isolationdocsSecret obfuscationDisabled-by-default outbound secret masking, environment and file discovery, reversible placeholders, and configurationdocs
Grok BuildCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Git worktree
Persistent memory
Apache-2.0 first-party code
Grok Build overviewTUI, headless mode, ACP, custom models, installation, and authenticationdocsHeadless and scriptingOne-shot execution, JSON event streams, sessions, ACP, update control, and CI integrationdocsPermissions and sandboxOS isolation profiles, default-off behavior, filesystem boundaries, and network limitsdocsPermissionsAsk, auto, always-approve, accept-edits and dont-ask modes plus allow and deny rulesdocsWorktreesParallel checkout creation, base refs, subagent isolation, cleanup, and Git limitationsdocsSubagents and extensionsDefault-on subagents plus skills, plugins, hooks, MCP, and LSP extensionsdocsSubagentsBuilt-in and custom child agents, parallel tasks, permissions, models, and worktree isolationdocsHooksTool and session lifecycle scripts, trust, blocking semantics, plugins, and execution scopedocsModes and commandsPlan review, session branching, rewind, background tasks, and control modesdocsPlan modePlan-only edit gate, approval transition, persisted plan, and permission-mode interactiondocsBackground tasksRecurring loops, background agents, task list, scheduling, interruption, and persistencedocsAgent dashboardMulti-session monitoring, agent status, attention requests, attachment, and controlsdocsEnterprise deploymentManaged policy, sandbox pinning, authentication, network, telemetry, ZDR, and bypass lockdocsSettings referenceModels, permissions, sandbox, MCP, subagents, memory, telemetry, and configuration precedencedocsGrok Build launchSubscription access, plan approval, diffs, and parallel subagentsannouncementOpen-source announcementSource release, extension system, MCP, subagents, and local inference configurationannouncementGrok Build changelogBeta status and current 0.2.111 product release verificationannouncementGrok Build repositoryPinned public source sync, runtime, tests, session architecture, sync policy, and licenserepositorySession and rewind guideLocal session storage, resume, fork, rewind, compaction checkpoints, and recovery boundariesrepositoryMemory guideOptional cross-session memory, default state, storage, retrieval, and lifecyclerepositoryMonitoring and OpenTelemetry guidePinned session telemetry, OpenTelemetry configuration, local defaults, exported signals, and operational privacy boundariesrepositoryMCP serversProject and user MCP configuration, HTTP and stdio transports, OAuth, server scopes, tool approval, and imported compatibility filesdocsSettings overviewTUI and config.toml settings, user and project scopes, starter configuration, precedence, and links to detailed referencesdocsCLI referenceGrok Build commands and common flags for prompts, sessions, headless operation, permissions, sandboxing, models, and updatesdocs
AiderCoding harness; Pair programmer
Coding harness
Pair programmer
Single-agent loop
Host-first
No first-party isolation
Session-based
Apache-2.0
Aider documentationProduct scope, terminal workflow, repository editing model, and documentation indexdocsOptions referenceComplete CLI flags including confirmation, automation, Git, tests, analytics, watch mode, and browser UIdocsModel connectionsCloud providers, OpenAI-compatible APIs, and local modelsdocsGitHub CopilotUse of models billed through a Copilot subscriptiondocsLocal models with OllamaLocal inference configuration, model warnings, context size, and environment requirementsdocsRepository mapRepository-wide symbol contextdocsGit integration and undoAutomatic commits, diffs, history, and rollbackdocsLinting and testingAutomatic linting, configured test commands, post-edit verification, and interactive test executiondocsScripting AiderOne-shot headless execution, shell scripting, confirmation bypass, and unsupported Python API boundarydocsAider in an IDEEditor-agnostic watch-files integration, AI comments, and terminal continuation workflowdocsExperimental browser UILocal Streamlit browser interface, experimental status, direct file edits, and Git commitsdocsOptional web scraping supportThe /web scraping command, optional Playwright dependency, and distinction from browser automationdocsAnalytics controlsOptional PostHog telemetry, local logging, collected fields, and disable controlsdocsAider benchmark methodologyProject-owned model benchmark tasks, edit formats, execution method, and interpretation limitsdocsAider v0.86.2 sourcePinned stable source, test and workflow inventory, benchmark assets, license, and contributor materialrepositoryPrivacy policyService and programming-tool data collection, randomly generated analytics identifiers, service providers, security limits, and opt-out controlsdocsRelease historyDated project release notes through v0.86.1 and the documented release-history boundary relative to the newer pinned source commitannouncementOperational FAQSupported workflows, provider data flow, repository handling, analytics references, usage limits, and operational troubleshootingdocs
OpenHandsCoding harness; Agent platform
Coding harness
Agent platform
Multi-agent runtime
Sandbox-first
Container, Managed sandbox
Persistent memory
MIT core; separate enterprise terms
OpenHands V1 documentation indexSeparation of product and SDK surfaces plus the current first-party documentation inventorydocsOpenHands 1.11.0 sourcePinned OSS application source, licensing boundary, tests, CI workflows, and contributor documentationrepositorySandbox overviewDocker, unisolated local process, and remote execution choices in the V1 productdocsDocker sandboxRecommended container isolation, mounts, networking, image selection, and local prerequisitesdocsProcess sandboxHost-process execution, absence of container isolation, trusted-task warning, and platform limitsdocsHeadless modeNon-interactive execution, always-approve behavior, output modes, scripting, and CI usedocsSecurity and action confirmationConfirmation policies, action risk analyzers, user review flow, and direct-tool bypass boundarydocsTask Tool SetSynchronous delegation to specialized subagents, task isolation, results, and limitationsdocsBrowser useBrowser tool configuration, navigation and interaction capabilities, screenshots, and sandbox relationshipdocsModel Context ProtocolMCP server discovery, tool integration, configuration, authentication, and execution modeldocsConversation persistenceConversation state storage, restoration, resume workflows, event persistence, and storage backendsdocsPersistent memoryOpt-in cross-conversation two-tier memory, retrieval, storage, lifecycle, and privacy considerationsdocsObservability and tracingOpenTelemetry spans, supported backends, instrumentation, debugging, and export configurationdocsMetrics trackingToken, cost, and latency metrics plus conversation-level aggregation and exportdocsLLM subscriptionsChatGPT Plus and Pro authentication for Codex models and subscription constraintsdocsLocal LLMsLocal-model configuration, GPU guidance, model requirements, and documented functional limitationsdocsEvaluation harnessProject-owned evaluation infrastructure and protocol surface without importing a product scoredocsCLI installationSupported CLI installation methods, local requirements, startup, configuration paths, upgrades, and platform setupdocsCLI MCP server managementHTTP, SSE, and stdio MCP server configuration, OAuth and header authentication, enablement, environment variables, and CLI managementdocsScheduled automationsScheduled unattended conversations, fresh sandbox creation, stored secrets, provider credentials, MCP integrations, run history, and user scopedocsEvent-based automationsGitHub-event and custom-webhook triggers, organization routing, filters, signing-secret workflow, service accounts, and failure modesdocsCustom sandbox imagesV1 agent-server container boundary, custom image construction, local-GUI scope, fallback behavior, base images, and runtime configurationdocsOpenHands 1.11.0 releaseDated OSS product release corresponding to the pinned implementation snapshot, distinct from separately versioned cloud releasesannouncement
gooseCoding harness; General-purpose agent
Coding harness
General-purpose agent
Delegated subagents
Host-first
OS sandbox
Session-based
Apache-2.0
Supported LLM providersHosted providers, local models, subscription authentication, and provider-specific setupdocsSubagentsIndependent delegated agents, automatic task delegation, and parallel workdocsHeadless gooseNon-interactive goose run usage for scripts, servers, batch work, and CIdocsSession recipesReusable workflows, structured output, retries, shell checks, and validation stepsdocsDeveloper extensionDefault autonomous host access, command and file tools, approval modes, and per-tool controlsdocsSession managementSQLite-backed sessions, resume, duplication, import, export, and cross-interface continuitydocsSmart context managementAutomatic context compaction, summarization, truncation, clearing, and prompt controlsdocsgoose v1.25.0 sandboxSeatbelt filesystem and network sandboxing for goose Desktop on macOSannouncementComputer Controller extensionBuilt-in MCP extension for browser, application, and desktop interactionsdocsCodebase analysisRepository structure, semantic focus, call graphs, delegation, and the 50,000-character output guarddocsLogging systemLocal session, tool-call, server, model, and token-usage logsdocsHarbor evaluation toolingProject-owned Terminal-Bench tooling and one-attempt snapshots; no benchmark score is importedrepositorySecurity guideFirst-party security hub for Adversary Mode, prompt-injection detection, classifier deployment, and MCP safety guidancedocsAdversary ModeOptional independent tool-call reviewer, plain-language policy, default tool scope, fail-open behavior, and block decisionsdocsPrompt-injection detectionOptional pattern and ML detection, thresholds, user alerts, full-permission execution, privacy boundary, and stated limitsdocsClassification API specificationSelf-hosted prompt-injection classifier contract, response labels, confidence handling, and sensitive-content transmission warningdocsSecurity policyOfficial vulnerability reporting, disclosure, and supported-version policy for the goose projectrepositorygoose v1.44.0 releaseCurrent release pin with security fixes, manual approval enforcement, scoped approvals, classifier hardening, and provider updatesannouncementGHSA-r5pp-p5r8-466rHigh-severity arbitrary command execution in goose review before 1.44.0 and the first patched releaseannouncement
ClineCoding harness; Coding agent
Coding harness
Coding agent
Multi-agent runtime
Host-first
Git worktree
Session-based
Apache-2.0
Cline overviewIDE, CLI, SDK, API, browser use, Kanban, enterprise, provider, and automation surfacesdocsProvider catalogFirst-party catalog of supported hosted providers and their shared configuration flowdocsLocal modelsOllama, LM Studio, and local inference configuration with hardware and model guidancedocsCLI overviewInteractive terminal, JSON output, headless runs, and automationdocsCLI referenceDefault auto-approve behavior, session resume, command policies, scheduling, MCP, and runtime flagsdocsAuto Approve and YOLOGranular IDE approvals, risk controls, YOLO mode, command allowlists, and unattended execution warningsdocsCheckpointsFile snapshots, comparison, and rollbackdocsKanbanResearch-preview parallel agents, per-task Git worktrees, review, dependency chains, and shipping workflowdocsSubagentsParallel read-only research agents and their limitsdocsAgent TeamsPersistent team state, coordinator delegation, shared task board, and inter-agent mailboxdocsMCPLocal and remote MCP server configuration, tools, resources, transport, and trust considerationsdocsSchedulingCron schedules, background hub execution, task management, logs, and unattended agent behaviordocsCline SDKEmbeddable TypeScript runtime, core packages, tools, sessions, plugins, and production integrationsdocsPermission handlingSDK permission decisions, automatic and explicit approval paths, callbacks, and policy implementationdocsPlugin and hook modelLifecycle hook stages, blocking and asynchronous policies, fail-open or fail-closed behavior, tool-call gates, observability, and plugin executiondocsPlugin installationExecutable plugin sources, npm and Git dependency installation, global and project scopes, extension points, and current surface limitationsdocsEnterprise MCP controlsOrganization MCP blocking, local allowlists, managed remote servers, personal-server restrictions, and policy precedencedocsEnterprise YOLO controlOrganization-wide disablement of unattended YOLO mode, policy precedence, monitoring expectations, and security tradeoffsdocsMemory BankUser-maintained project documentation for cross-session continuity and its manual workflowdocsOpenTelemetry integrationOptional OTLP log and metric export, configuration, backends, and data-governance controlsdocsCline v4.0.11 sourcePinned source, tests, workflows, security policy, contributor guide, and evaluation framework staterepositoryCline evaluation frameworkProject-owned smoke and E2E tasks, trials and metrics, disabled CI layers, and unfinished nightly evaluationrepository
Gemini CLICoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Container, Git worktree
Persistent memory
Apache-2.0
Gemini CLI consumer transitionConsumer cutoff date, Antigravity migration, retained enterprise access, API-key access, and ongoing Gemini CLI supportannouncementGemini CLI v0.52.0 sourcePinned stable source, license, tests, workflows, project eval assets, and contributor materialrepositoryStable release 0.52.0Stable version, release date, installation channel, feature highlights, and upgrade guidanceannouncementSandboxingOpt-in Seatbelt, Docker, Podman, gVisor, LXC, Windows isolation, profiles, mounts, and limitsdocsTrusted foldersDefault-off folder trust, untrusted-workspace restrictions, project configuration, MCP, and auto-accept boundariesdocsHeadless modeNon-interactive prompts, text and structured outputs, workspace trust, scripting, and CI behaviordocsSubagents and browser agentBuilt-in and custom subagents, disabled-by-default browser agent, Chrome modes, consent, policies, and sandbox interactiondocsCheckpointingDefault-off shadow-Git snapshots, conversation and tool-call capture, restore workflow, and requirementsdocsPersistent context and memoryGEMINI.md hierarchy, durable memory files, explicit save, inspection, reload, and context guidancedocsExperimental Auto MemoryDefault-off transcript mining, review inbox, candidate memory and skills, data flow, privacy, and limitationsdocsMCP integrationMCP server configuration, discovery, transports, trust, authentication, tool filtering, and managementdocsTelemetryDefault-off OpenTelemetry logs, metrics, traces, local and Google Cloud exporters, and data fieldsdocsGemini CLI evaluation assetsProject-owned behavioral and safety evaluations without a complete independent product result recordrepositoryPolicy engineAllow, deny, and ask decisions, rule priority, headless behavior, admin policies, and the disabled Workspace tierdocsLifecycle hooksSession, agent, model, and tool lifecycle hooks, deterministic blocking, configuration scopes, and security implicationsdocsExperimental Git worktreesOpt-in isolated checkouts for sessions and subagents, lifecycle, branch handling, and filesystem-only boundariesdocsEnterprise admin controlsAdministrator-enforced authentication, extensions, MCP, sandboxing, and approval-mode restrictionsdocs
Antigravity CLICoding harness; Coding agent
Coding harness
Coding agent
Multi-agent runtime
Host-first
OS sandbox
Session-based
Proprietary binary distribution
Antigravity CLI overviewShared agent harness, terminal and SSH surface, headless focus, settings sync, conversation export, and Gemini CLI migrationdocsGemini CLI consumer transitionConsumer migration, enterprise continuity, launch scope, shared harness direction, and retained Gemini CLI mechanismsannouncementPlans and quotasIndividual and enterprise access paths, plan-dependent quotas, CLI availability, third-party models, and changing limitsdocsModel availabilityPlan-specific managed Gemini, Claude, and GPT-OSS model availability without implying harness performancedocsInstallation and authenticationNative installation, agy command, Google sign-in, secure credential storage, SSH flow, and enterprise onboardingdocsExecution modesRequest-review, accept-edits, planning, permission bypass, review boundaries, and subagent inheritancedocsFine-grained permissionsDeny, ask, and allow precedence, file and command policy, browser and MCP actions, workspace defaults, and promptsdocsNative terminal sandboxOpt-in nsjail, sandbox-exec, and AppContainer isolation, default state, interactive containment, and one-command escapedocsBackground tasks and subagentsParallel asynchronous agents, custom agent definitions, task monitoring, reasoning logs, tool outputs, and approval routingdocsConversation managementWorkspace-scoped history, session resume, conversation forks, and the explicit absence of filesystem branchingdocsPlugins and skillsPlugin packaging, skills, custom subagents, hooks, rules, workspace scope, and MCP configurationdocsMCP integrationLocal and remote MCP servers, transports, configuration scope, authentication, discovery, and permission promptsdocsCLI settings referenceCommands, headless controls, review presets, sandbox and workspace defaults, telemetry, rewind, and observability panelsdocsAntigravity CLI 1.1.8 releasePinned shipped version, typed NDJSON traces, tool and subagent trajectory fields, JSON-schema output, token accounting, and public support-repository boundaryrepositoryArtifact reviewPlan and diff artifacts, approval and rejection, line-level feedback, media review, and pre-write co-steeringdocsCLI projectsProject-scoped conversations, startup selection, project creation, resume behavior, and cross-project conversation forksdocs
GitHub Copilot CLICoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Managed sandbox
Persistent memory
GitHub Copilot CLI License
About GitHub Copilot CLITerminal workflow, Copilot plans, MCP, custom agents, memory, custom providers, and local modelsdocsProgrammatic referenceHeadless execution, model pinning, tool and URL permissions, secrets redaction, and transcriptsdocsCloud and local sandboxesFilesystem, network, and system isolation plus preview status and platform limitationsdocsFleet subagentsTask decomposition, parallel subagents, model selection, context isolation, and orchestrationdocsSession rollbackConversation rewind, Git snapshots, tool-based file restore, constraints, and destructive effectsdocsCopilot MemoryPublic-preview repository facts, user preferences, CLI use, citations, validation, enablement, retention, and governancedocsUsing Copilot CLIInteractive modes, approval defaults, tool execution, sessions, autonomy, and day-to-day workflowdocsLocal and cloud sandboxesPreview status, local OS isolation, hosted cloud environments, boundaries, billing, and platform constraintsdocsCustom providersOpenAI-compatible, Anthropic, Azure, and local Ollama provider configuration and limitationsdocsMCP configurationBuilt-in GitHub MCP, additional servers, transports, tools, authentication, trust, and managementdocsAutopilot modeAutonomous continuation, programmatic use, permission interaction, explicit continuation limits, and stop conditionsdocsCopilot hooksPre-tool policy decisions, lifecycle events, subagent completion, audit logging, validation, and security cautionsdocsCLI customization overviewCustom instructions, hooks, skills, custom subagents, MCP, plugins, and personal settingsdocsCustom agents and subagentsAgent profiles, scoped tools and MCP, built-in agents, separate contexts, delegation, and reusable specializationdocsHooks referenceLifecycle event schemas, pre-tool and permission decisions, stop controls, matcher filtering, outputs, and failure behaviordocsCopilot CLI v1.0.75 support repositoryPinned installer and changelog support tree, release workflows, distribution license, and absent core sourcerepository
Cursor CLICoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Git worktree
Session-based
Proprietary
Cursor CLI overviewInteractive agent, model selection, rules, MCP, installation, subscription access, and product scopedocsHeadless modePrint mode, structured output, scripting, non-interactive file changes, trust, force behavior, and automationdocsParameters and isolation controlsSandbox, worktree, trust, yolo, MCP, session resume, headless output, ACP, and worker parametersdocsCLI permissionsShell, read, write, web-fetch, and MCP allowlists and denylists plus project and global scopesdocsCLI configurationApproval modes, sandbox and network settings, rewind, workspace trust, web search, and config precedencedocsCLI changelogDated sandbox, worktree, subagent, checkpoint, rewind, MCP, trust, and auto-review changes through July 2026announcementAgent Client ProtocolEditor integration, session protocol, model and mode selection, tools, MCP, and streaming eventsdocsGitHub ActionsCI authentication, headless execution, secrets, permissions, event triggers, and review automationdocsCursor Agent CLI announcementRelationship to Cursor Agent, initial headless scope, subscription models, approvals, and risk noticeannouncementCursor public support repository snapshotPinned README, security policy, and issue template surface; it does not expose Cursor CLI source, tests, releases, or evaluationsrepositoryCLI installationSupported macOS, Linux, and Windows installation paths, version verification, upgrades, uninstallation, and PATH setupdocsUsing Agent in the CLIInteractive prompting, rules, MCP use, conversation navigation, change review, command approval, and session workflowsdocsCLI MCPModel Context Protocol server configuration, tool discovery, connection management, and CLI invocationdocsCLI Shell ModeContinuous shell interaction, retained conversational context, command history, and long-running development sessionsdocsCLI authenticationInteractive login and API-key authentication used by local, headless, and continuous-integration CLI workflowsdocs
Junie CLICoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
Git worktree
Session-based
Proprietary
Junie CLI quick startTerminal installation, authentication, model selection, approval modes, and core workflowdocsModel selectionJetBrains subscription, BYOK providers, custom profiles, proxies, model aliases, and effort levelsdocsOllama integrationLocal model discovery, OpenAI-compatible endpoints, manual profiles, and capability caveatsdocsCustom subagentsAutomatic delegation, isolated contexts, tool restrictions, models, skills, and EAP statusdocsParallel sessions and worktreesConcurrent sessions, file-conflict risks, Git worktree isolation, and change transferdocsHeadless modeCI execution, authentication, project trust, restricted mode, and configuration loadingdocsAction AllowlistDefault ask behavior, sensitive actions, read-only commands, command patterns, MCP approval, and allowlist storagedocsMCP configurationLocal and remote MCP servers, OAuth, configuration scope, commands, tools, and authenticationdocsGuidelines and memoryAGENTS.md and Junie guideline scopes, loading precedence, persistent instructions, and context limitsdocsRemote modeSynchronized browser access to a live local CLI session, supported interactions, subscription requirements, and terminal-only limitationsdocsAgent SkillsOpen Agent Skills format, progressive loading, project and user scopes, automatic invocation, and cross-agent importsdocsJunie CLI configurationUser and project configuration, path precedence, secure project-trust markers, restricted loading, and non-interactive rollout limitsdocsJunie CLI hooksEAP lifecycle hooks, shell execution, pre-tool allow, ask, or deny decisions, timeouts, scopes, and security boundariesdocsAgent Client ProtocolACP editor integration, local and remote agent setups, project context, protocol startup, and authentication surfacedocsAudited Junie 2518.1 support repositoryPinned release installers, channel registries, templates, small distribution test surface, and proprietary licenserepositoryAudited Junie nightly 2518.1 releaseImmutable proprietary nightly distribution release used as the repository audit boundary rather than a stable-channel claimannouncement
Factory DroidCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Git worktree
Session-based
Proprietary
Droid CLI referenceInteractive and headless modes, output formats, permissions, worktrees, sessions, conversation rewind, MCP, and CIdocsDroid ExecRead-only default, autonomy tiers, structured JSON-RPC, model selection, parallel worktrees, and sessionsdocsBring Your Own KeyHosted providers, custom endpoints, open-source models, Ollama, vLLM, and local executiondocsCustom DroidsDelegated subagents, isolated contexts, prompt and model selection, and per-agent tool policiesdocsIDE integrationsVS Code and JetBrains plugins, diff viewing, editor selection context, and shared diagnosticsdocsOS sandboxBeta opt-in OS isolation, per-command and whole-process modes, defaults, fail-closed behavior, coverage, and limitsdocsAutonomy levelsOff, low, medium, and high approval postures, allow and block rules, organization caps, and riskdocsDroid ControlPlugin-provided browser, terminal, Electron, and desktop control plus verification and demo workflowsdocsSession memory managementExplicit memory files, AGENTS.md, compaction, sessions, context practices, and cross-session limitationsdocsTelemetry exportEnterprise OpenTelemetry export, OTLP configuration, events, attributes, security, and observability scopedocsFactory v0.180.0 release notesCurrent binary version, dated feature and hardening changes, compatibility, and product-surface distinctionsannouncementFactory public support repositoryPinned documentation and distribution support tree, workflows, benchmark pages, proprietary notice, and absent core sourcerepositoryTerminal-Bench methodologyFive-run benchmark method, agent integration, sandbox assumption, model comparisons, and limitationsannouncementDroid settingsUser and project settings hierarchy, local overrides, autonomy defaults, command policy, cloud session sync, hooks, MCP timeouts, and mission settingsdocsDroid MCP configurationInteractive and scripted MCP management, registry, HTTP, SSE, and stdio transports, OAuth, credentials, package pinning, and tool trustdocsDroid hooksLifecycle hooks, pre-tool blocking, subagent and session events, deterministic automation, credential scope, and trust risksdocsLLM safety and agent controlsModel-as-untrusted framing, command policy, secret and DLP controls, programmable enforcement, sandbox layers, and admin governancedocs
ForgeCodeCoding harness; Extensible harness
Coding harness
Extensible harness
Delegated subagents
Host-first
Git worktree
Session-based
Apache-2.0
Installation and setupPlain-language product identity, installation, subscriptions, cloud and local models, and provider logindocsAudited ForgeCode readmeInteractive, one-shot, and Zsh modes; conversations, agents, skills, providers, MCP, and semantic searchrepositoryLatest verified releaseRelease recency, signed release commit, current provider fixes, and versioned distribution artifactsannouncementApache-2.0 licenseRoot source-code license at the audited repository commitrepositoryCLI implementationInteractive default, one-shot prompt and stdin, conversation resume and export, logs, agents, and worktree flagrepositoryZsh workflowOptional shell integration, prompt routing, agent switching, conversation commands, and diagnosticsdocsCustom and local providersSelf-hosted endpoints, enterprise gateways, localhost examples, provider overrides, and session selectiondocsBuilt-in provider registryForty-seven audited provider definitions including subscriptions, hosted APIs, local runtimes, and compatible endpointsrepositoryMCP integrationOptional project and user MCP servers, transports, trust, OAuth, CLI management, and external browser examplesdocsReusable skillsBuilt-in and custom SKILL.md workflows, loading precedence, resources, discovery, and invocationdocsCustom agentsAgent definitions, model selection, tool restrictions, MCP tool globs, and agent-as-tool supportdocsBuilt-in agent rolesForge, Muse, and Sage roles, write access, task selection, context continuity, and version-control guidancedocsDelegated task toolSubprocess agents, isolated task context, concurrent delegation, agent selection, results, and resume identifiersrepositoryBuilt-in implementation agentBroad host tools, MCP and task access, code-edit workflow, tool-assisted compilation, tests, and verification instructionsrepositoryPermission policyRestricted-mode prerequisite, allow, deny and confirm rules, allow-all generated policy, exemptions, and MCP bypassdocsConfiguration defaultsRestricted mode off by default, tool and request limits, retries, semantic search, HTTP, and compaction settingsdocsGenerated permission defaultsAllow-all defaults for built-in reads, writes, commands, and network fetches when the policy file is generatedrepositoryGit worktree implementationThe sandbox-named option creates or reuses a Git worktree and branch without an operating-system boundaryrepositoryForgeCode ServicesOptional workspace login, sync, semantic search, stored source chunks and embeddings, ignore rules, and deletiondocsService privacy policySource-file and embedding storage for semantic search, direct model-provider routing, and service data handlingdocsFile snapshots and undoPer-path content snapshots used for file-level undo rather than multi-file workspace or conversation checkpointsrepositoryLocal log commandsListing, tailing, and selecting local ForgeCode log files for operational inspectionrepositoryProject evaluation frameworkFirst-party task runner, parallelism, timeouts, data-driven cases, validations, and debug artifacts; not independent evidencerepositoryContinuous integration workflowCross-platform build and test automation, coverage, performance checks, and release jobs at the audited commitrepository
Qwen CodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Container, Git worktree
Persistent memory
Apache-2.0
Qwen Code 0.21.0 source snapshotImmutable Apache-2.0 source snapshot for the audited stable release, including CLI, desktop, tests, workflows, and one project-owned codegraph eval fixturerepositoryQwen Code 0.21.0 releaseStable release identity and dated version boundary used for the auditannouncementSandboxOpt-in macOS Seatbelt, Docker and Podman isolation; provider selection; filesystem and network profiles; and the network-open macOS defaultdocsApproval and settings referenceDefault approval for edits and shell, plan, auto-edit, classifier-based auto mode, and unrestricted yolo modedocsHeadless safety and budgetsStructured automation, explicit run budgets, and the warning that yolo neither enables nor substitutes for a sandboxdocsPersistent memoryQWEN.md instructions, default-on auto-memory, user and project scopes, review commands, cleanup, and optional Git synchronizationdocsCheckpointingDisabled-by-default shadow-Git snapshots, local conversation capture, restore behavior, and configurationdocsCommands and recoverySession resume and branching, conversation rewind, checkpoint restore, exports, budgets, and background-agent commandsdocsBuilt-in Computer Use releaseZero-configuration desktop Computer Use as a native capability rather than an inferred browser feature from WebFetch or MCPannouncementModel providersOpenAI-compatible, Anthropic, Gemini, Qwen, Vertex AI, and local self-hosted model configurationdocsOpenTelemetry observabilityDefault-off local or OTLP logs, metrics, spans, tool and subagent events, sensitive-data controls, and outbound trace propagation boundariesdocsQwen Code overviewOpen-source agent scope, terminal workflow, supported integrations, installation entry points, and official documentation mapdocsMCP server integrationModel Context Protocol server configuration, transport and tool discovery, lifecycle management, and extension pointsdocsQwen Code ExtensionsInstallable extension packaging for agents, skills, MCP servers, context files, workspace scopes, enablement, and removaldocsqwen serve daemonAlpha local HTTP and event-stream API, Web Shell, sessions, persistence, loopback and remote authentication, TLS, and maturity limitsdocsJetBrains integrationJetBrains IDE connection through the Agent Client Protocol, setup, session interaction, and editor context exchangedocsNested subagent releaseDated release notes for nested subagent spawning, default depth controls, tree visibility, parameter-level permissions, and Web Shell sessionsannouncement
Continue CLICoding harness; Coding agent
Coding harness
Coding agent
Single-agent loop
Host-first
No first-party isolation
Session-based
Apache-2.0
Continue CLI quickstartTerminal agent, interactive and headless modes, authentication, configuration, models, MCP, and automation flagsdocsTool permissionsAllow, ask, and exclude policies, safe headless defaults, persistent permissions, and CLI overridesdocsModels, rules, and toolsReusable local and shared models, behavioral rules, MCP tools, and configuration compositiondocsFinal Continue source snapshotImmutable final Apache-2.0 source snapshot and explicit notice that the repository is read-only and no longer actively maintainedrepository
Mistral VibeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
Git worktree
Session-based
Apache-2.0
Pinned Mistral Vibe source treeAudited v2.22.0 CLI implementation, package layout, tests, documentation, and release-aligned source staterepositoryPinned Apache-2.0 licenseLicense text for the open CLI codebase, kept separate from model and hosted-service termsrepositoryMistral Vibe v2.22.0 releaseVersioned release provenance for the exact source commit used by this auditrepositoryInstall and setupSupported operating systems, Python requirement, Mistral-plan access, API keys, offline operation, and setup flowdocsWork with the CLIInteractive and programmatic modes, tool review, trust, structured output, budgets, sessions, and cost-estimate caveatdocsSafety, approvals, and permissionsLayered agent policy, trusted-folder boundary, per-tool rules, outside-directory confirmation, and auto-approve riskdocsAgents and subagentsBuilt-in approval profiles, plan mode, delegated explore agent, custom subagents, and AGENTS.md discoverydocsLifecycle hooksPre-tool, post-tool, and post-agent policy hooks, strict failure behavior, structured payloads, and subagent inheritancedocsMCP serversMCP transports, static authentication, tool naming, permissions, and the documentation's stale OAuth limitationdocsSkillsReusable skill format, discovery paths, enable and disable rules, and user-invocable commandsdocsOffline and local modelsvLLM, llama.cpp, LM Studio, Ollama, generic OpenAI-compatible providers, hardware guidance, and offline network controlsdocsCLI configurationProvider presets, OpenRouter example, tools, hooks, configuration precedence, and default-on telemetry and updatesdocsCLI and editor surfacesCLI, VS Code, ACP-compatible editors, and the boundary between local execution and remote web sessionsdocsVibe Code Web sandbox boundarySeparate single-tenant web sandbox, outbound-network posture, credential scope, ephemerality, and missing network allowlistsdocsPinned shipped defaultsDefault provider and model catalog, local llama.cpp provider, compaction, policy bypass, telemetry, OTEL, connectors, and session settingsrepositoryPinned shell permission implementationHost-shell execution, ask-by-default posture, safe-command allowlist, denylists, sudo sensitivity, timeout, and output caprepositoryPinned built-in agent definitionsDefault, plan, accept-edits, auto-approve, explore-subagent, and Lean profiles with their actual overridesrepositoryPinned CLI entrypointCurrent programmatic approval behavior, explicit auto-approve, budgets, structured output, worktrees, trust, and resume flagsrepositoryPinned checkpoint and rewind managerMessage-level rewind, recorded file states, restoration selection, in-place history truncation, and restore-error reportingrepositoryPinned Git worktree implementationOptional per-task branch and worktree creation, reuse, cleanup, and the limits of workspace rather than process isolationrepositoryPinned local-session architectureDurable local session format, append-friendly messages, atomic metadata, resume, rewind, and compatibility constraintsrepositoryPinned CI workflowPre-commit, CLI startup checks, pytest retry policy, and separately enforced snapshot testsrepositoryPinned GitHub ActionFirst-party composite action for installing and invoking programmatic Vibe in CIrepositoryPinned changelogVersion history for approval-default changes, rewind, subagents, MCP OAuth, hooks, OTEL, sessions, and provider supportrepositoryVibe product evolutionOfficial distinction and handoff among terminal, editor, and separately managed remote Vibe surfacesannouncementVibe overviewCoding-product scope, supervised local workflows, remote sandbox option, and common user tasksdocsCLIArena Mistral Vibe adapters and Terminal-Bench runsIndependent, inspectable discovery signal only. Reported Terminal-Bench 2.0 runs use a modified Mistral Vibe fork and do not yet meet HarnessMatch's full immutable configuration and replication requirements, so no score is imported.discovery
Kimi CodeCoding harness; Coding agent
Coding harness
Coding agent
Multi-agent runtime
Host-first
No first-party isolation
Session-based
MIT
Kimi Code 0.29.2 source snapshotImmutable MIT source snapshot for the audited release, including 1,256 test-like files and eight workflows but no dedicated product-evaluation suiterepositoryKimi Code 0.29.2 releaseStable package release and dated version boundary used for the auditannouncementKimi Code documentationSkills, hooks, parallel subagents, MCP, installation, TUI, and extension pointsdocsProviders and modelsKimi, Anthropic, OpenAI-compatible, Responses API, Gemini, Vertex, and custom provider routingdocsCommand referenceStructured print mode, permission modes, ACP integration, authenticated local web service, session visualizer, session selection, and output formatsdocsConfiguration defaultsManual interactive permissions, default-on anonymous telemetry, background-task behavior, subagent timeouts, and effectively unbounded print-mode ceilingsdocsBuilt-in tools and swarmsAgent and AgentSwarm approval behavior, inherited permissions, background execution, timeout defaults, and configurable swarm concurrencydocsSessions and contextStructured local Wire event streams, resume, replay, fork, subagent trace history, and export without claiming learned cross-session memorydocsHooksLifecycle events, pre-tool blocking, observability hooks, timeouts, and explicit fail-open semanticsdocsAgent SkillsProject, user, extra, and built-in skill scopes, automatic or direct invocation, progressive loading, scripts, and cross-agent directoriesdocsAgents and subagentsBuilt-in and custom agents, parallel isolated contexts, background execution, permission inheritance, tool restrictions, and local trace storagedocsModel Context ProtocolProject and user MCP scopes, local process trust boundary, tool permissions, OAuth, approval behavior, and YOLO-mode warningdocsPluginsOfficial and third-party marketplaces, trust labels, Git and ZIP installation, plugin skills, MCP servers, managed copies, and user-global scopedocsData locationsLocal configuration, credentials, session history, logs, skills, plugin state, relocation, and cleanup boundariesdocsEnvironment controlsData-root relocation, telemetry disablement, background-task lifetime and concurrency limits, model overrides, and diagnostic controlsdocs
Letta HarnessCoding harness; Extensible harness
Coding harness
Extensible harness
Multi-agent runtime
Host-first
Managed sandbox
Persistent memory
Apache-2.0
Letta Code 0.29.4 source snapshotImmutable Apache-2.0 client and harness source snapshot, including 580 test-like files and eleven workflows; hosted Letta services remain outside the auditrepositoryLetta Code 0.29.4 releaseStable release identity and dated version boundary used for the auditannouncementLetta Code quickstartCLI, desktop and web surfaces, headless mode, coding-plan access, persistent memory, parallel conversations, and schedulesdocsMemFSGit-backed long-term memory, versioned memory edits, system context, skills, synchronization, and shared memory repositoriesdocsPermissionsUnrestricted default, standard approvals, edit-only mode, persistent allow and deny rules, tool restriction, and memory guardsdocsExecution environmentsLocal host execution, managed cloud sandboxes, remote machines, environment-specific files and credentials, and memory separationdocsCloud sandboxesManaged isolated computers, shell and filesystem scope, lifecycle, local-file separation, and agent-state boundariesdocsSubagentsForeground and background delegation, isolated contexts, persistent-agent delegation, custom tools, skills, and memory scopedocsAgent skillsAgent, project, computer, and bundled skill scopes plus persistence, installation, trust warnings, and direct invocationdocsHeadless modeNon-interactive execution, structured streams, permission control, environment routing, persistent conversations, and tool eventsdocsSchedulesDurable cloud and local schedules, execution targets, cloud-sandbox fallback, limits, timezones, and run historydocsSupported model-provider typesHosted, BYOK, subscription OAuth, Ollama, LM Studio, vLLM, SGLang, OpenRouter, and other provider types exposed by the Letta platformdocsMemory research lineageFirst-party MemGPT paper that motivates tiered persistent memory; research lineage rather than comparative product-performance evidenceannouncementSleep-time compute researchFirst-party research behind offline memory refinement, used as mechanism evidence rather than a harness benchmarkannouncementCurrent Letta documentation indexCurrent product map and naming boundary for the Letta Harness, agent SDK, API, execution environments, memory, automation, and customizationdocsMCP tool execution modelExternal MCP server registration, transports, authentication, remote execution boundary, attachment to agents, and guidance to prefer skills in the app and CLIdocsClient tool execution modelLocal client execution, approval handoff, server-versus-client boundaries, local credentials and resources, and Letta Harness as reference implementationdocsPinned CLI MCP implementationCode-verifiable HTTP, SSE, and stdio MCP server commands, headers, authentication tokens, argument parsing, and CLI result handling at v0.29.4repositoryGitHub ActionRepository installation, issue and pull-request triggers, @letta-code mentions, workflow setup, authentication, and automated coding or review tasksdocsTrusted runtime modsFully trusted in-process extensions for tools, commands, hooks, permissions, providers, UI, reload behavior, sharing, and explicit trust boundarydocsHarness configurationHierarchical global and project settings, backend and authentication modes, provider configuration, environment selection, permissions, and defaultsdocsHarness changelogVersioned Letta Harness changes, permission defaults, memory, channels, mods, providers, schedules, and documentation-lag warningannouncementCLI referenceInteractive and headless commands, backend selection, model and memory controls, permission modes, session resume, environment routing, and automation flagsdocs
Kilo CodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
OS sandbox, Managed sandbox, Git worktree
Session-based
MIT
Audited Kilo Code source treePublic monorepo implementation inspected at an immutable commitrepositoryMIT licenseClient and agent source license plus OpenCode lineage attributionrepositoryKilo Code v7.4.16 releaseLatest verified release, dated artifacts, CLI binaries, VSIX packages, and indexing packagerepositoryInstallation guideVS Code, Open VSX, CLI, JetBrains, and supported installation channelsdocsKilo CLITerminal workflow, autonomous execution, permission configuration, session continuation, logs, and OpenTelemetry exportdocsCloud AgentBrowser interface, managed Linux container, repository branches, remote control, triggers, and beta limitsdocsBrowser useBuilt-in VS Code browser automation and optional Playwright MCP path in the CLIdocsAI providersBuilt-in Gateway, BYOK providers, provider allowlists, and local or self-hosted pathsdocsChatGPT subscription accessOAuth use of eligible ChatGPT subscriptions, supported surfaces, and cloud exclusionsdocsOllama local modelsLocal Ollama configuration, hardware caveats, context setup, and offline model pathdocsMCP configurationLocal and remote MCP transports, OAuth, project and global configuration, and tool permissionsdocsCustom subagentsBuilt-in and custom delegated agents, isolated sessions, Task invocation, concurrency, and per-agent permissionsdocsAgent permissionsAllow, ask, and deny policy rules, precedence, sensitive files, and subagent delegation controlsdocsDefault approval behaviorBroad out-of-box tool allowances, shell and sensitive-file prompts, and runtime auto-approval riskdocsOS sandboxingOpt-in macOS and Linux confinement, network policy, write boundaries, unsupported Windows, and trusted-integration limitsdocsCheckpoint recoveryDefault-on Git snapshots, message-level rollback, per-file CLI recovery, exclusions, and retention caveatsdocsCodebase indexingOpt-in Tree-sitter chunking, embeddings, local and hosted vector stores, filtering, and semantic searchdocsSessions and sharingPrivate resumable sessions, read-only sharing, forking, and retained task contextdocsAgent ManagerParallel VS Code sessions, Git worktree isolation, review workflow, terminals, and approval routingdocsBenchmarking statusSeparation of current smoke-eval evidence from unverified Harbor, ATIF, Opik, and comparison roadmap itemsdocsRelease smoke evaluation workflowTwo vendor-operated Harbor smoke tasks, fixed model, private kilo-bench dependency, artifact upload, and cost comments; not an independent benchmarkrepositoryContinuous integration workflowCross-platform unit matrix, sandbox setup, HTTP API exerciser, JetBrains tests, and required gatesrepositorySandbox implementation defaultCode-verifiable disabled default, network deny default, destination validation, and project-policy tighteningrepositoryRepository security policyDisclosure channel and stale pre-sandbox threat-model language, retained as a documented source contradictionrepositoryLocal sandbox threat modelOS confinement design, unrestricted read scope, write and network controls, supported platforms, and explicit privacy-boundary and firewall limitationsannouncementKilo CLI product surfaceCLI sandbox command, workspace write confinement, read-only Git metadata, optional network denial, sessions, providers, and automation surfacedocsKilo Cloud security architectureManaged Cloudflare sandbox containers, repository credentials, trust boundaries, persistence, observability, and cloud-agent isolation architecturedocsOpenRouter coding appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
Command CodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
Git worktree
Persistent memory
Proprietary
CLI referenceInteractive and print modes, sessions, worktrees, model selection, permissions, and IDE setupdocsCustom agentsParallel subagents, isolated context, tool policy, background execution, and model selectiondocsPermissionsDefault read-only access, approval-gated writes and shell, project trust, local path scope, optional telemetry, and headless safety behaviordocsMCP integrationMCP server configuration, tool exposure, authentication, and permission integrationdocsHeadless modeRead-only automation default, explicit yolo mutation mode, ten-turn default, exit codes, transcript persistence, and deterministic resume flagsdocsFile checkpointsAutomatic per-message backups, file and conversation restore modes, 10 MB exclusion, per-session scope, and all-or-nothing restore validationdocsHooksPre-tool policy decisions, stop checks, session context injection, audit hooks, retry caps, execution semantics, and failure behaviordocsAgent skillsProgressively loaded project and user skills, direct invocation, trust boundaries, and reusable workflow instructionsdocsPublic support repository snapshotImmutable public support and product-information snapshot; it does not expose the proprietary agent implementation, tests, or evaluation suiterepositoryGit worktreesSession isolation in Git worktrees through the slash command, CLI flag, and agent-controlled enter and exit toolsdocsSessions and checkpointsDurable JSONL transcripts, resume, fork, clone, scratchpad, sharing, export, compaction, checkpointing, and rewind behaviordocsGoal modeStanding multi-turn objectives, autonomous continuation, verification loop, progress handling, and explicit goal cancellationdocsMemory instructionsAGENTS.md instruction tiers, project and user scopes, path imports, precedence, and assembly into agent requestsdocsTaste profilesLearned preference profiles, enable and disable controls, review, sharing, management commands, and persistence boundariesdocsSettings and configurationUser and project JSON configuration, environment variables, MCP and keybindings, scopes, precedence, and documented defaultsdocsPlan modeReasoning-before-execution mode, plan review, inline comments, revision, approval, and auto-accept behaviordocsInteractive modeTerminal session surface, input modes, keyboard controls, editor handoff, model switching, and user interaction defaultsdocsBackground tasks and schedulingBackground commands, monitors, scheduled wakeups, task ledger, background subagents, cron scheduling, sleep, and TUI managementdocsCommand Code v1 changesVersioned summary of the rewritten permission engine, worktrees, sessions, background tasks, context handling, and extension changesannouncementMods extension APITypeScript extension API for tools, commands, lifecycle hooks, event observers, rendering, model providers, packaging, and verificationdocsCommand Code StudioHosted dashboard, Taste Studio, usage and billing, API keys, organization administration, and account control surfacedocsOpenRouter coding appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
CodebuffCoding harness; Extensible harness
Coding harness
Extensible harness
Multi-agent runtime
Host-first
No first-party isolation
Session-based
Apache-2.0 code with hosted service
Agent overviewDelegation, built-in agents, context passing, review, research, and deterministic orchestrationdocsQuick startCLI installation, project workflow, authentication, and interactive usagedocsSDK and programmatic accessTypeScript SDK, programmatic agent calls, event stream, session continuation, and automationdocsLocal chat history and troubleshootingLocal conversation history, client version checks, installation recovery, and terminal automation guidancedocsHow Codebuff worksOrchestrator and subagent pipeline, repository code map, local execution, tests, and transmitted contextdocsDefault autonomy and context managementNo-prompt host execution, parallel editing, automatic review, context compaction, and project-owned evaluationsdocsExecution modesDefault, Lite, Max, and no-write Plan modes plus review, typecheck, and test behaviordocsKnowledge and project instructionsRepository knowledge files, scoped instructions, and project verification commandsdocsLarge-project guidanceVendor-documented large-repository workflow, scoped working directories, and knowledge organizationdocsLicense, subscription, privacy, and isolation FAQApache-2.0 client, Codebuff subscription, data handling, host access, and optional Docker isolationdocsCustom agent definitionsTypeScript agents, model selection, tools, nested subagents, and deterministic control flowdocsPinned public implementationOpen-source implementation, OpenRouter model support, CLI, agent framework, SDK, browser use, and test instructionsrepositoryPinned browser agentBrowser automation agent backed by Chrome DevTools MCP tools and structured browser resultsrepositoryPinned MCP clientMCP client transports, connection setup, tool discovery, tool calls, and error handlingrepositoryPinned agent runtime notesAgent definitions, generator isolation, tool access, subagents, and the distinction from repository command isolationrepositoryPinned BuffBench frameworkProject-owned git reconstruction tasks, AI judging, trace analysis, cost tracking, and evaluation workflowrepositoryOpenRouter coding appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
CrushCoding harness; Pair programmer
Coding harness
Pair programmer
Delegated subagents
Host-first
No first-party isolation
Session-based
FSL-1.1-MIT
Pinned Crush overviewTerminal workflow, providers, Hyper, local models, sessions, LSP, MCP, skills, hooks, permissions, configuration trust, logs, and storagerepositoryCrush configuration schemaProvider endpoints, model configuration, permissions, tools, MCP, and session settingsdocsCrush v0.87.0 releaseLatest reviewed release, MCP OAuth, Channels preview, LSP improvements, event reconnection, and signed artifactsrepositoryHyper provider plansFirst-party free credit tier, paid subscription, hosted coding models, and zero-data-retention claimdocsPinned CLI root commandInteractive terminal entry point, session continuation, yolo mode, project logging, and run command examplesrepositoryPinned non-interactive run commandPrompt input from arguments or stdin, quiet and verbose output, model selection, session reuse, and event streamingrepositoryPinned experimental server commandExperimental shared server and client mode for multiple terminal clientsrepositoryPinned task-agent implementationDefault task-agent construction, child sessions, parallel delegation, and agent-tool invocationrepositoryPinned task-agent instructionsRead-only glob, grep, listing, and file-view tools exposed to the delegated research agentrepositoryPinned agent and provider configurationCoder and task agents, read-only task permissions, provider setup, MCP configuration, tools, logging, and agent settingsrepositoryPinned permission managerPer-tool prompts, allow once, session grants, denials, pre-approved tools, hook decisions, and automatic approval moderepositoryPinned hooks documentationPreToolUse allow, deny, and rewrite behavior plus the documented lack of interception inside subagentsrepositoryPinned session storeSQLite-backed sessions, child task sessions, messages, usage, cost, and todo persistencerepositoryPinned log commandProject log inspection, follow mode, and runtime diagnostic outputrepositoryPinned MCP lifecycleMCP server lifecycle, tools, prompts, resources, OAuth handling, and channel integrationrepositoryPinned contributor guideProject architecture, coder and task agents, SQLite, LSP, MCP, test commands, and contribution workflowrepositoryPinned licenseFSL-1.1-MIT terms, competitive-use restriction, and future MIT license conversionrepositoryPinned cross-platform CI workflowUbuntu, macOS, and Windows builds plus race-enabled Go test executionrepositoryPinned agent testsProject-owned VCR-backed tests for agent tool use, including reading, writing, and shell executionrepositoryPinned server end-to-end testsProject-owned end-to-end tests for server events, agent turns, cancellation, and session behaviorrepositoryOpenRouter coding appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
MuxCoding harness; Agent platform
Coding harness
Agent platform
Multi-agent runtime
Host-first
Git worktree, Container
Persistent memory
AGPL-3.0
Mux 0.28.1 source snapshotImmutable AGPL-3.0 source snapshot with application code, 879 test-like files, thirteen workflows, persistent-memory implementation, and project-owned Terminal-Bench adaptersrepositoryMux 0.28.1 releaseStable release identity and dated version boundary used for the auditannouncementMux agentsAgent definitions, nested delegation, tool policy, model defaults, internal desktop automation agent, and background memory-consolidation agentdocsMux workspacesParallel workspaces, task separation, lifecycle, Git integration, and workspace statedocsMux runtimesLocal, worktree, SSH, Docker, dev-container, and provisioned runtime boundaries, including the lack of isolation in local modedocsModel providersAnthropic, OpenAI, Google, xAI, Moonshot, OpenRouter, Bedrock, GitHub Copilot subscription access, Ollama, and custom local OpenAI-compatible endpointsdocsMux MCP serversMCP configuration and per-workspace processes; browser automation appears as an external Chrome MCP example rather than a native featuredocsAdministrative policy fileFail-closed startup, last-known-good refresh behavior, and restrictions over providers, models, MCP servers, and runtimesdocsCLI and run budgetsHeadless goal runs, structured events, turn and spending limits, runtime selection, and automation exit conditionsdocsTelemetryDefault anonymous product events, excluded data, transparent source payload, and opt-out environment variabledocsTerminal-Bench toolingProject-owned benchmark adapter, runtime and provider inputs, run artifacts, and leaderboard-submission workflow; not independent comparative evidencedocsLocal runtimeDirect execution in the user's working copy, shared files, parallel conflict warning, and explicit absence of isolationdocsDocker runtimePer-workspace container isolation, Git-bundle synchronization, command execution, image configuration, and teardowndocsDev Container runtimeProject-defined container execution, host worktree creation, tool prerequisites, command boundary, and cleanupdocsWorktree runtimeParallel file isolation, shared Git metadata, branch visibility, host execution, layout, and instruction-based constraintsdocsSSH runtimeRemote tool execution, hostile-host threat model, default credential non-forwarding, explicit project secrets, and sync boundarydocsCoder runtimeCoder workspace discovery and provisioning, SSH connection path, template requirements, and shared-host behaviordocsTool hooksExperimental pre- and post-tool scripts for command blocking, linting, validation, environment setup, and policy decisionsdocsInit hooksRepository setup script execution during workspace creation, executable requirements, environment, and failure behaviordocsGitHub ActionsCI use of mux run, provider and GitHub secrets, exit-code merge gating, runtime selection, and workflow examplesdocsACP editor integrationsStdio Agent Client Protocol bridge for Zed, Neovim, and JetBrains with sessions, tool delegation, and streamingdocsVS Code extensionPreview editor pairing for VS Code and Cursor across local and SSH workspaces, commands, and chat surfacedocsInstruction filesShared and Mux-specific AGENTS.md layering, model prompts, project and global scopes, precedence, and migration caveatsdocsPlan modeRead-oriented planning, plan-file-only edits, explicit review, external edits, diff detection, and execution handoffdocsInstallationSigned macOS, Linux, and alpha Windows packages, development builds, platform support, and update channelsdocsServer accessDefault bearer-token authentication, token resolution, GitHub owner allowlist, sessions, network access, and no-auth warningdocsOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
Coder AgentsCoding harness; Agent platform
Coding harness
Agent platform
Multi-agent runtime
Managed-first
Managed sandbox, Container
Session-based
AGPL-3.0 core with separately licensed enterprise components
Coder Agents overviewBeta status, self-hosted control-plane loop, web chat, workspaces, subagents, persistence, and built-in toolsdocsCoder Agents architectureControl-plane loop, workspace execution path, tool boundaries, stored chat state, identity, and network isolation conditionsdocsModels and providersAdministrator-selected providers, OpenAI-compatible and self-hosted endpoints, model options, routing, credentials, and BYOKdocsBuilt-in toolsVisible tool calls, user-scoped permissions, workspace creation, file operations, command execution, and tool limitsdocsPlatform controlsAdministrative model and prompt policy, template allowlists, MCP policy, lifecycle, retention, and spend controlsdocsMCP server controlsAdministrator-configured MCP servers, authentication, availability policies, tool allowlists, denylists, and identity headersdocsWorkspace skills and MCPRepository skills and workspace MCP discovery, transports, timeouts, tool naming, and path-safety rulesdocsCoder Agents getting startedVersion prerequisite, roles, setup, API automation, template selection, network caveat, human review, and beta limitationsdocsChats REST APIExperimental programmatic chat sessions, child-agent records, message state, model configuration, MCP selection, and streamingdocsSpend managementPer-user and per-model cost, token, message, chat, and rolling spend-limit recordsdocsChat debug loggingOptional structured traces of requests, responses, tool activity, token use, retries, errors, export, and retentiondocsVirtual desktopExperimental computer-use subagent, desktop interaction, provider requirements, workspace module, and API configurationdocsChat runtime architecture at inspected commitDatabase state machine, queueing, retries, worker ownership, API surface, streaming, and durable chat transitionsrepositorySubagent implementation at inspected commitChild-chat creation, model override resolution, subagent messaging, interruption, and agent listing implementationrepositoryWorkspace MCP implementation at inspected commitWorkspace MCP discovery, provider-safe tool names, connection routing, parallel tool metadata, and error handlingrepositoryWorkspace execution implementation at inspected commitHeadless shell execution, timeouts, background process tracking, working directories, environment, and output boundsrepositoryCoder Agents runtime tests at inspected commitFirst-party automated tests for chat runtime, providers, MCP, permissions, streaming, retries, and workspace behaviorrepositoryCoder repository licenses at inspected commitAGPL-3.0 license for the open-source repository outside separately licensed enterprise directoriesrepositoryCoder enterprise license at inspected commitSeparate license terms for code under the repository enterprise directoryrepositoryCoder repository overview at inspected commitSelf-hosted platform scope, workspace infrastructure, native agent loop, model routing, identity, cost tracking, and audit claimsrepositoryOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
Zoo CodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
Git worktree
Session-based
Apache-2.0
Zoo Code product pageVS Code positioning, modes, diff review, MCP, model-agnostic access, indexing claims, and the headless claim requiring qualificationdocsZoo Code documentation overviewCurrent extension surface, local filesystem and terminal workflow, modes, orchestrator, and product-scope framingdocsZoo Code model providersProvider profiles, hosted APIs, gateways, local endpoints, and the breadth of model accessdocsOpenAI and ChatGPT accessDirect OpenAI API support and separate ChatGPT Plus or Pro subscription sign-indocsUsing local modelsLocal and offline model operation through Ollama, LM Studio, and compatible endpointsdocsModes and orchestratorMode-specific tools, sticky model selection, custom modes, and Orchestrator delegation through new_taskdocsTool workflowRead, edit, command, workflow tools, explicit extension approvals, and the new_task delegation surfacedocsMCP integrationGlobal and project MCP configuration, external servers, per-tool allow controls, and optional browser serversdocsCodebase indexingTree-sitter parsing, embedding providers, Qdrant storage, incremental updates, multi-folder behavior, and privacy boundariesdocsCheckpointsDefault-on shadow-Git snapshots, task-scoped file restoration, diffs, and the lack of pre-command checkpointsdocsGit worktreesSeparate branches and VS Code windows for user-managed parallel work, plus worktree requirements and limitationsdocsFile access rules.rooignore enforcement for file tools and its explicit limitation as neither full command coverage nor a system sandboxdocsMessage queueingSequential queued prompts and the important behavior that a queued message implicitly approves the next pending actiondocsDiagnostic exportVersioned error details, full task history, actions, environment, configuration, and support-oriented diagnostic exportdocsRoo to Zoo migrationProject succession, configuration migration, compatibility boundaries, and user transitiondocsZoo Code 3.72.0 releaseLatest stable extension version, publication date, VSIX artifact, provider changes, and subtask fixesannouncementRepository overview at inspected commitActive Roo succession, VS Code extension scope, current feature summary, contribution path, and release lineagerepositoryApache license at inspected commitApache-2.0 terms for the reviewed source repositoryrepositorySecurity policy at inspected commitSupported-version scope, private vulnerability reporting, acknowledgment target, and remediation targetrepositoryContribution guide at inspected commitIssue-first contribution policy, testing requirements, and the stated future goal of establishing harness evaluationsrepositoryCode QA workflow at inspected commitCross-platform unit and integration tests, coverage, lint, type checks, dependency review, and security-oriented source checksrepositoryDelegation implementation at inspected commitApproval-gated hierarchical child creation and replacement of the parent by one sole active child taskrepositoryAuto-approval policy at inspected commitGranular read, write, command, MCP, mode, subtask, and follow-up auto-approval decisionsrepositoryWorktree service at inspected commitCode-level Git worktree creation, listing, switching, deletion, branch handling, and host-process executionrepositorySource CLI readme at inspected commitPrint and streaming JSON modes, source-build instructions, default auto-approval, and stale Roo-owned installation pathsrepositorySource CLI manifest at inspected commitPrivate Roo-branded package metadata and version, limiting its status as a distributed Zoo product interfacerepositorySource CLI behavior at inspected commitNon-interactive execution, provider selection, persistence, and automatic approval unless approval mode is requestedrepositoryOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
ZCodeCoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Managed-first
Container
Session-based
Proprietary
ZCode Agent workflowDesktop-native task workflow, file and browser context, execution modes, Git state, model switching, and review flow without importing GLM model claimsdocsGoal ModeSession-level goal state, resource budgets, automatic per-round verification, failed-closed completion, pause and resume, recovery, and trajectory recordsdocsSubagentsBuilt-in and custom subagents, independent contexts, tool scoping, foreground parallel execution, and current Beta limitsdocsMCP serversMCP service connection, tool availability, configuration, and workspace usedocsSafety confirmationFive permission modes, exact action prompts, scoped allow and reject policy, risk visibility, review flow, and per-file open and undo actionsdocsADE toolsTask workspace, existing Docker and SSH targets, built-in browser preview and element context, terminal, DevTools, and review integrationdocsRemote developmentExecution placement in SSH hosts and existing local Docker containers, connection logs, target-account boundaries, and reconnection behaviordocsHooksJSON subprocess protocol, session and tool events, pre-tool policy decisions, post-tool diagnostics, stop checks, scopes, limits, and local-code warningdocsTask and file managementTask status and grouping, repository file search, Git change filtering, review context, and generated repository orientationdocsPlugin systemBeta marketplace packages spanning skills, commands, subagents, MCP, hooks, and language servers plus enable and configuration behaviordocsAgent SkillsSKILL.md format, user and project scopes, explicit invocation, enablement, generation, and imports from other coding agentsdocsModel connectionsZ.ai subscriptions, first-party and OpenRouter API routes, Anthropic and OpenAI compatible custom endpoints, and self-hosted service supportdocsUsage statisticsDevice-local session, token, message, activity, and model-use records plus remote Z.ai plan and tool-call quota reportingdocsRemote ControlMobile access to a desktop-open workspace, task continuation, connection dependency, and remote-control limitsdocsBot ChannelWeChat and Feishu task entry, workspace selection, task continuation, and desktop service dependencydocsDesktop installationVersion 3.5.2 desktop distribution for macOS and Windows, Linux beta status, onboarding, and workspace startupdocsRelease 3.5.2 and recent automation changesVersion and release date, scheduled tasks, background subagents and shell tasks, built-in web integration, and session recovery fixesdocsOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
stagewiseCoding harness; Agent platform
Coding harness
Agent platform
Multi-agent runtime
Host-first
Git worktree
Session-based
AGPL-3.0
stagewise product overviewCurrent product positioning, desktop IDE workflow, provider choice, parallel agents, and the unsupported cache-hit marketing claimdocsInstall stagewiseNative Electron desktop distribution for macOS, Windows, and Linux rather than a browser-hosted productdocsHow agents workAgent loop, file and shell tools, approval behavior, and independent agents running concurrently with separate chats and modelsdocsBrowser and agentEmbedded Chromium, DOM context, screenshots, navigation, tabs, page interaction, and browser-aware codingdocsDiff reviewFull edit history, inline diffs, per-hunk accept and reject controls, tool-call undo, redo, and design previewsdocsAgent contextWORKSPACE.md, skills, optional AGENTS.md, mentions, DOM context, context compression, and environment updatesdocsModels and providersStagewise Account models, imported provider subscriptions, direct API keys, custom endpoints, and local modelsdocsCustom providersOpenAI-compatible, Responses, Anthropic, Azure, Bedrock, Vertex, Ollama, LM Studio, vLLM, and self-hosted endpointsdocsBYOK setupDirect provider credentials and imported Claude, GPT, Gemini, Kimi, Qwen, DeepSeek, GLM, and MiniMax accessdocsWorkspacesMultiple project folders, Git worktree choices, and the explicit warning that connected workspaces are not sandboxeddocsSkills and pluginsSKILL.md extensions and built-in product plugins, distinct from the unconfirmed Model Context Protocol capabilitydocsStagewise 1.25.0 releaseLatest stable version, publication date, packaged desktop artifacts, and current release provenancerepositoryRepository overview at inspected commitOpen-source desktop architecture, development status, supported platforms, and repository organization at the inspected refrepositoryAGPL license at inspected commitAGPL-3.0 licensing terms for the reviewed open-source repositoryrepositorySecurity policy at inspected commitPrivate vulnerability reporting process and the repository security-policy artifactrepositoryContribution policy at inspected commitContributor documentation and the restriction against external feature or code pull requestsrepositoryMonorepo CI at inspected commitLint, build, unit-test, cross-platform, and Electron packaging checks in the public repositoryrepositoryTool approval modes at inspected commitAlways-ask, always-allow, and smart approval modes, including always-ask as the persisted defaultrepositoryHost shell tool at inspected commitShell command execution through the local host shell service rather than an operating-system sandboxrepositoryJavaScript worker at inspected commitElectron utility-process and node:vm boundary used for the restricted JavaScript execution toolrepositoryRestricted filesystem wrapper at inspected commitMount-scoped filesystem checks for the JavaScript worker, not isolation for ordinary shell executionrepositoryDiff history service at inspected commitCode-level diff history primitives supporting review, attribution, undo, and redo of agent editsrepositoryProduct chat agent at inspected commitThe single registered chat-agent type and its absent finish schema, limiting productized child-agent delegationrepositoryGeneric child-agent primitive at inspected commitLow-level synchronous and asynchronous child-agent plumbing that is not exposed as a registered product toolrepositoryInternal smoke-test CLI at inspected commitPrivate minimal Anthropic-only headless smoke utility and its forced always-allow approval posturerepositoryInternal CLI package manifest at inspected commitPrivate version-zero package metadata showing that the smoke CLI is not a public product interfacerepositoryRepository evaluation asset at inspected commitThe narrow copywriting-skill evaluation file, which is not a coding-harness benchmark or independent product evaluationrepositoryOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
Hermes AgentCoding harness; General-purpose agent
Coding harness
General-purpose agent
Multi-agent runtime
Host-first
Container, Managed sandbox
Persistent memory
MIT
Hermes Agent 0.19.0 releaseLatest stable version and date, desktop and gateway changes, smart approvals, subagent inspection, delivery durability, and completion-contract release contextrepositoryHermes documentation overviewSupported installation surfaces, persistent memory, terminal backends, scheduled automation, delegation, browser automation, MCP, and batch processingdocsTools and toolsetsCoding file and terminal tools, browser automation, delegation, scheduling, memory, and per-platform toolset configurationdocsSecurity and trust boundariesSmart, manual, and off approval modes; immutable deny rules; fail-closed timeouts; file-write policy; SSRF controls; container boundaries; and host-execution limitationsdocsTerminal backend configurationLocal, Docker, SSH, Singularity, Modal, and Daytona execution placement, container hardening and persistence, credential forwarding, and remote file synchronizationdocsCheckpoint and rollbackOpt-in pre-mutation shadow-Git snapshots, file and directory rollback, diff preview, retention limits, excluded files, and external-side-effect limitsdocsSubagent delegationParallel and nested children, inherited tool boundaries, separate contexts and terminals, iteration budgets, background delivery, cancellation, and restart-durability limitsdocsPersistent goalsDurable goal lifecycle, continuation budget, completion contracts, evidence-oriented judging, persistence, intervention, and documented false-positive and false-negative risksdocsScheduled tasksGateway scheduler, isolated sessions, execution ledger states, overlap locking, delivery targets, pause and resume, and restart handlingdocsAPI serverAuthenticated OpenAI-compatible endpoints, structured streaming, background run surfaces, and jobs CRUD for scheduled automationdocsDesktop appElectron desktop client, headless JSON-RPC backend, local and remote runtime attachment, authentication, logs, and extension modeldocsPersistent memoryBounded cross-session memory, FTS5 session search, default autonomous writes, optional write approval, threat scanning, and review controlsdocsMCP integrationMCP server configuration, transport and tool loading, authorization, filtering, cache refresh, and security considerationsdocsBatch processingParallel prompt execution, isolated environments, resumable runs, and structured trajectory generation for training or evaluation without a product scoredocsRepository snapshotPublic feature and setup surface at the exact inspected commit, including interfaces, providers, memory, cron, delegation, runtime backends, and coding entry pointsrepositoryPackage version at inspected commitHermes Agent package identity, version 0.19.0, Python constraints, dependencies, and executable entry pointsrepositorySecurity policy at inspected commitExplicit OS-boundary threat model, host default, terminal-backend versus whole-process isolation, in-process heuristic limits, credentials, plugins, and external surfacesrepositoryContinuous integration at inspected commitChanged-area detection and parallel test, lint, type, integration, desktop, gateway, security, and packaging lanes in the public repositoryrepositoryCheckpoint implementation at inspected commitShadow-Git checkpoint capture, restore, diff, pruning, size bounds, excluded paths, project identity, and non-fatal failure behaviorrepositoryApproval implementation at inspected commitDangerous-command detection, deny and allow rules, session and permanent decisions, cron behavior, and backend-dependent guard pathsrepositoryDelegation implementation at inspected commitParallel and nested child execution, concurrency limits, inherited tools, asynchronous delivery, cancellation, live logs, and lifecycle handlingrepositoryAutomation API at inspected commitOpenAI-compatible requests, structured run events, authentication, session continuity, background runs, and scheduled-job endpointsrepositoryObservability specification at inspected commitTrace and metric contracts for model, tool, approval, delegation, lifecycle, correlation, and exporter behavior plus observer limitationsrepositoryProject SWE runner at inspected commitProject-owned local, Docker, and Modal task runner and trajectory format, recorded only as an evaluation asset without importing a scorerepositoryBatch trajectory runner at inspected commitProject-owned concurrent agent trajectory generation and checkpointed batch execution, not independent harness performance evidencerepositoryOpenRouter appsDiscovery signal only; usage rank is not used as a quality or capability score.discovery
mini-SWE-agentCoding harness; Coding agent
Coding harness
Coding agent
Single-agent loop
Host-first
Container, OS sandbox
Session-based
MIT
mini-SWE-agent 2.4.6 releaseCurrent audited release, published from the inspected commitrepositoryRepository readme at inspected commitProject scope, installation, intentionally minimal design, models, environments, and evaluation workflowsrepositoryPackage metadata at inspected commitVersion 2.4.6, Python requirements, package status, dependencies, entry points, and MIT license metadatarepositoryCLI modesInteractive terminal usage, confirm, yolo, and human modes, interruption controls, and local historydocsControl flowLinear agent loop, bash-only action surface, termination, and exception handlingdocsExecution environmentsHost execution, Docker, Singularity, Bubblewrap, ConTree, and managed SWE-ReX environmentsdocsModel quickstartLiteLLM-backed provider configuration, API keys, model names, and gateway examplesdocsLocal model configurationLocal OpenAI-compatible or vLLM endpoints, custom model names, and manual cost registrationdocsTrajectory output formatVersioned trajectory records, full configuration, model statistics and calls, messages, exit status, and submission outputdocsTrajectory inspectorLocal browser-based inspection of saved trajectory filesdocsSWE-bench runnerBatch task selection, parallel workers, environment options, output, and interrupted-run behaviordocsProgramBench runnerProgramBench task selection, parallel execution, environment configuration, and outputdocsYAML configurationExplicit agent, model, environment, and runner configuration for reproducible runsdocsPython bindingsProgrammatic setup and invocation outside the interactive terminal UIdocsContribution guideDevelopment installation, testing, documentation, and contribution workflowdocsSecurity reporting at inspected commitProject security reporting contacts and disclosure channelrepositoryTest workflow at inspected commitProject-owned automated test workflow across supported Python versionsrepositoryInteractive confirmation implementationPer-action confirmation, yolo switching, human input, and interruption behaviorrepositoryDefault agent implementationSingle linear agent loop, bash command parsing, step limits, and trajectory serializationrepositoryHost environment implementationDirect local subprocess execution with inherited host accessrepositoryExperimental Bubblewrap environmentOptional Linux namespace and filesystem isolation implementationrepositorySWE-bench runner at inspected commitProject-owned batch runner, parallel workers, completed-trajectory handling, and result extractionrepositoryProgramBench runner at inspected commitProject-owned ProgramBench batch and trajectory generation utilitiesrepository
AmpCoding harness; Extensible harness
Coding harness
Extensible harness
Delegated subagents
Host-first
Managed sandbox
Session-based
Proprietary
Amp owner’s manualCurrent product surface, terminal and editor interfaces, thread model, modes, built-in tools, and supported platformsdocsProjects and changes workflowRepository-linked settings and secrets plus ship, branch, pull-request, and custom changes workflowsdocsCloud orbsManaged cloud execution environment, lifecycle, setup and resume scripts, portals, secrets, and billingdocsRemote runnersLocal or remote runner registration and dispatch to user-controlled machinesdocsSubagents and reviewAutomatic isolated-context subagents, delegation limits, and separate review agentsdocsMCP and workspace trustLocal and remote MCP, skill-scoped tools, workspace approval, OAuth, loading order, and server policydocsPermissions and pluginsNo-approval default and optional allow, reject, ask, or delegated rules through plugins and managed settingsdocsScheduled agentsOne-time and recurring schedules that wake a thread with its saved prompt, context, and historydocsThread sharing and governancePrivate, group, workspace, and unlisted visibility plus administrator access boundariesdocsCLI execute modeInteractive CLI, non-interactive execute mode, orb execution, streaming JSON, and remote runner modedocsPlugin APITool and lifecycle hooks, custom agents and threads, executor selection, background work, and permission decisionsdocsTypeScript SDKProgrammatic local and orb execution, streams, structured options, permission rules, and orb-setting limitationsdocsSecurity referenceService architecture, retention, thread audit trail, prompt-injection controls, secret redaction, and enterprise controlsdocsDurable agent foundationDistributed durable execution and unified monitoring across web, mobile, editor, and CLIannouncementAgents in Orbs announcementFresh managed machines, remote inspection, change synchronization, and non-interactive orb launchannouncementRemote runners announcementRemote thread creation and headless runner mode on user-controlled machinesannouncementSchedules announcementRecurring agent wakeups with saved prompt, context, history, and downstream integrationsannouncementCustom agents announcementPlugin-defined main agents, subagents, tool pipelines, background threads, and parallel custom workersannouncementSubscription and orb limitsSubscription tiers, included orb hours, mode access, linked model subscriptions, and pay-as-you-go continuationannouncement
Kiro CLICoding harness; Coding agent
Coding harness
Coding agent
Delegated subagents
Host-first
No first-party isolation
Session-based
Proprietary
Kiro CLI 2.14.0 releaseCurrent release, interrupted-response retries, and opt-in V2-to-V3 agent migrationannouncementCLI quickstartTerminal workflow, plan mode, approval prompts, project steering, and MCP onboardingdocsSubagentsUp to four parallel agents, task graphs, review loops, live monitoring, aggregation, and scoped tool permissionsdocsTool permissionsPer-tool trust, command and filesystem approvals, custom-agent permissions, and headless trust behaviordocsHeadless modeAPI-key authentication, non-interactive execution, scoped tool trust, MCP startup gates, CI examples, and limitationsdocsCLI commands and sessionsInteractive and non-interactive flags, session resume, JSON-formatted settings, diagnostics, and log controlsdocsSession managementPer-turn autosave, per-directory resume, export and import, SQLite storage, and custom persistence scriptsdocsConversation rewindNon-destructive conversation branching with tool-call and context previews; not file rollbackdocsGoal loopIterative implementation and agent-driven verification, acceptance criteria, iteration limits, steering, and cancellationdocsQueue steeringMid-turn redirection at tool boundaries, queued follow-ups, cancellation, and timing limitationsdocsClassic checkpointingOpt-in shadow-Git snapshots, per-turn and per-tool diffs, soft and hard restore, session scope, and cleanupdocsBuilt-in toolsFile, shell, search, goal, web, subagent, and other built-in tool semantics and settingsdocsHooksLifecycle and tool hooks, JSON events, blocking pre-tool decisions, logging, formatting, and validationdocsMCP integrationUser and workspace MCP servers, transport, OAuth, tool loading, trust, and enterprise registry controlsdocsCustom agentsAgent-specific prompts, tools, resources, permissions, hooks, MCP servers, and configuration scopedocsSettings referenceCheckpoint feature flag, compaction, logs, tool search, Kiro home isolation, and version-specific settingsdocsCLI 3.0 early accessOpt-in unified harness, capability policies, specs, breaking changes, session incompatibility, and known gapsdocsCLI 2.7 goal and steering releaseIntroduction of goal loops, queue steering, and enriched rewind previewsannouncementData protectionEncryption, individual and enterprise service-improvement treatment, telemetry, and opt-out controlsdocsPrivacy and security modelLocal execution boundary and explicit distinction between Supervised or Autopilot review behavior and sandboxing, isolation, or access controldocsCLI 3.0 permission rulesEarly-access declarative allow, ask, and deny capability rules, precedence, matching, project scope, and auditabilitydocs
Poolside Agent CLICoding harness; Coding agent
Coding harness
Coding agent
Single-agent loop
Host-first
Container, Managed sandbox, Git worktree
Session-based
Proprietary client
Poolside Agent CLIInteractive, one-shot, and ACP interfaces plus session and approval workflowdocsCLI installation and authenticationInstaller, Poolside deployment and standalone login, environment credentials, configuration, logs, and trajectoriesdocsInteractive modeApproval modes, interruption, conversation-only rewind, session recovery, trajectory viewer, and alternate ACP serversdocsAutomated modeOne-shot prompts, JSONL output, continuation, exit codes, environment authentication, and unsafe auto-allowdocsCLI referenceCurrent commands and flags for TUI, exec, ACP, history, MCP, secrets, configuration, and updatesdocsACP editor integrationPoolside as an ACP server for compatible editors and session configurationdocsTool and path policy referenceScoped allow and deny rules, path permissions, secrets, MCP, Docker sandbox, and network policy settingsdocsManaged sandboxesDefault local environment, managed local and deployment-dependent remote sandboxes, role access, filesystem and egress boundaries, and remote-MCP exceptiondocsOrganization permissionsRBAC for agents, auto approval, credentials, MCP, repositories, sandboxes, and cross-user trajectoriesdocsManaged agentsReusable model, instruction, tool, repository, credential, skill, MCP, and sandbox configurations with runtime access intersectiondocsAgent skillsLocal, project, and organization-managed skills, access controls, and sandbox mount limitationsdocsIndexed repositoriesOrganization-managed repository indexing and agent access controls for broader code contextdocsPoolside model deploymentProprietary Poolside models deployed within the customer's environment; kept separate from harness capabilitydocsPoolside platform deploymentOn-premises and air-gapped platform deployment, models, control plane, storage, and infrastructure requirementsdocsPoolside CLI 1.0.13 releaseCurrent binary release and published distribution artifactsrepositoryRelease repository readme at inspected commitOpenRouter, Ollama and OpenAI-compatible local endpoints, ACP client and server roles, worktrees, MCP, settings, and permission syntaxrepositoryRelease changelog at inspected commitVersion history through 1.0.13, JSON logging, worktrees, session persistence, compaction, and provider additionsrepositoryDistribution license at inspected commitProprietary end-user license notice for the distributed clientrepository
PlandexCoding harness; Coding agent
Coding harness
Coding agent
Single-agent loop
Host-first
No first-party isolation
Persistent memory
MIT
Repository snapshot and lifecycle noticePinned product overview, cloud wind-down notice, self-hosting path, scripting interface, and Claude subscription supportrepositoryMIT licenseLicense text at the inspected repository commitrepositoryLast CLI release: v2.2.1Last published CLI release, dated 2025-07-16, and its installation artifactsrepositoryLocal-mode self-hostingDocker Compose server setup, local host selection, and provider credential flowrepositoryPinned install guideManual release installation, source build, and WSL-only Windows supportrepositoryCLI referenceScriptable commands, non-interactive flags, background-task constraints, rewind, and Claude subscription commandsrepositoryAutonomy levelsFive autonomy presets and the documented defaults for context, apply, execute, debug, and commit behaviorrepositoryAutonomy defaults in sourceCode-level default semi mode, disabled auto-apply and auto-exec, enabled auto-commit, and rewind configurationrepositoryPersistent plansStored plan context, conversations, pending changes, history, branches, and project-scoped plan organizationrepositoryContext managementProject maps, automatic context selection, smart context windows, manual loading, images, URLs, and piped inputrepositoryDiff review and applyPending-change sandbox, diff UI, per-file rejection, apply flow, rollback, and auto-apply behaviorrepositoryExecution and debuggingHost command generation, approval defaults, test and build loops, browser debugging, rollback, and destructive-command warningrepositoryVersion control and rewindPlan history, rewind, applied-file reversion, branches, and the warning that rewind history cannot itself be restoredrepositoryModel providersOpenRouter, direct provider precedence, OpenAI, Anthropic, Google, Azure, Bedrock, DeepSeek, Perplexity, and compatible endpointsrepositoryClaude Pro and Max subscriptionsSubscription connection, quota behavior, backup providers, and self-hosted supportrepositoryOllama local modelsSelf-host-only Ollama support, local model packs, hardware guidance, and explicit quality limitationsrepositoryModel rolesPlanner, architect, coder, builder, summarizer, and support roles used inside one orchestrated planrepositoryBuilt-in model packsRole-based provider combinations, local and hybrid packs, context limits, and intended pack tradeoffsrepositoryBackground tasksParallel background streams and their incompatibility with the default automatic context-loading moderepositorySecurity notesSensitive-file ignore behavior and ephemeral API-key handling; no repository vulnerability-reporting policy is providedrepositoryLinux process groupingBest-effort systemd user cgroup process cleanup and explicit fallback to no isolationrepositoryNon-Linux process groupingNo-op cgroup implementation on non-Linux operating systemsrepositoryBrowser debugging sourceVisible Chrome launch, console and JavaScript error capture, and the absence of a headless browser moderepositoryPromptfoo evaluation proof of conceptProject-owned prompt evaluation scaffold whose metrics section remains marked coming soonrepositoryDocker publishing workflowRelease-oriented image build and publishing automation rather than continuous test integrationrepository
Discovery watchlist
Products stay outside the ranked catalog until first-party evidence supports a complete coding-harness profile.
Terminus 2Terminal-Bench documents Terminus as an autonomy-first evaluation agent outside the task container. It remains a research scaffold until first-party availability and daily-use support are documented well enough for workflow recommendation.adjacent toolMiMo CodeThe official product page documents a CLI and context features, but detailed first-party permission, runtime, and recovery documentation is still limited.needs more evidenceRoo CodeThe first-party repository was archived in May 2026. Active lineage is represented by Zoo Code and archived tools are excluded from recommendation results.archived lineageSlate AgentDiscovered in ecosystem catalogs, but current first-party technical documentation is insufficient for source-backed capability classification.needs more evidenceOttiliDiscovered in ecosystem catalogs, but current first-party technical documentation is insufficient for source-backed capability classification.needs more evidenceAnteDiscovered in ecosystem catalogs, but current first-party technical documentation is insufficient for source-backed capability classification.needs more evidencePortkey and model gatewaysProvider gateways can be important harness dependencies, but they do not independently implement the coding-agent loop compared by HarnessMatch.adjacent toolGeneral personal agentsGeneral agents without a documented software-engineering execution profile remain outside the primary catalog until coding workflows are first-party documented.adjacent tool