Coding agent

GGCode

Terminal coding harness with policy modes, durable context, multi-agent execution, and governed worktrees.

Status: Active
License: MIT
Evidence: Documented + code-verifiable, 18 sources
Product record checked: 2026-08-01

At a glance

An MIT-licensed coding harness whose Go runtime implements an adaptive model-and-tool loop, repository and shell mutation, automatic context compaction, project memory, resumable JSONL sessions, provider switching, subagents, persistent teammates, independent worktree workers, policy modes, hooks, traces, checkpoints, MCP, and browser automation.

Good choice if

Terminal workflows that need several providers, local OpenAI-compatible endpoints, MCP, and built-in browser control
Long autonomous tasks that benefit from resumable context, parallel agents, scheduled execution, and local execution traces
Governed change delivery through isolated Git worktrees, configured checks, review, approval, promotion, and release stages

Check before choosing

Normal execution is host-first: path policy, protected-file rules, and Git worktrees constrain access and change organization but are not an OS or container security boundary
The sandboxed JavaScript VM is limited to programmatic calls to read-only tools; it does not isolate file writes, shell commands, MCP servers, plugins, hooks, browser actions, or external side effects
Bypass and autopilot modes deliberately weaken interactive approval, while harness workers launch independent GGCode processes in bypass mode inside an allowed worktree path

See 4 more considerations

Subagents and swarm teammates share the main process and workspace; only harness workers receive a separate process and worktree, and even that process is not an OS sandbox
File checkpoints cover edits made through GGCode tools and cannot reverse shell commands, Git operations, browser actions, messages, or external-service effects
Hooks and plugins execute trusted user-supplied code, MCP servers retain their own trust boundary, and the broad desktop, web, mobile, messaging, and automation surface increases configuration responsibility
Exported session traces and harness event logs improve local inspection, but they do not establish task success or independent product performance

Quick facts

Interfaces: Terminal, IDE, Web, Automation
Approval style: Policy rules
Model portability: Provider + local
Isolation: Git worktree
Agent organization: Multi-agent runtime
Local model path: DocumentedThe source establishes the mechanism, not its quality or availability in every mode.

View full technical record

Catalog layer: Coding harness
Harness membership: Qualifies: 4 required criteria evidenced
Product role: Coding agent
Agent organization: Multi-agent runtime
Runtime posture: Host-first
Isolation options: Git worktree
State model: Persistent memory
Context lifecycle: Persistent memory
Permission posture: Policy rules
Verification: Workflow-gated checks
Observability: Structured traces
Recovery: File checkpoint
Interfaces: Terminal, IDE, Web, Automation
Provider posture: Multi-provider
Model portability: Provider + local
Subscription access: Available
Enterprise access: Not documented
Local model path: DocumentedThe source establishes the mechanism, not its quality or availability in every mode.
Claim boundary: Harness capability only

Capability support

Documented first-class product support, checked against the sources below.

External tools (MCP): DocumentedProduct-supported MCP integrationSource · checked 2026-08-01The source establishes the mechanism, not its quality or availability in every mode.
Reusable skills: Not documentedNo first-class support established by the current recordAbsence of current documentation is not proof that the capability is impossible.
Local models: DocumentedLocal or self-hosted model pathSource · checked 2026-08-01The source establishes the mechanism, not its quality or availability in every mode.
Agent parallelism: DocumentedDelegated or parallel agent workflowSource · checked 2026-08-01The source establishes the mechanism, not its quality or availability in every mode.
Runs without an open UI: DocumentedNon-interactive or automation surfaceSource · checked 2026-08-01The source establishes the mechanism, not its quality or availability in every mode.
Browser control: DocumentedBuilt-in or product-supported browser controlSource · checked 2026-08-01The source establishes the mechanism, not its quality or availability in every mode.
Isolated execution: Not documentedNo first-class support established by the current recordAbsence of current documentation is not proof that the capability is impossible.
Undo file changes: DocumentedIn-process checkpoints for file edits made through GGCode toolsSource · checked 2026-08-01File checkpoints do not reverse shell, Git, browser, messaging, or external-service side effects.

Getting started

Install a stable GGCode build, use the onboarding wizard to select a provider, review the supervised default and protected-path rules, then enable bypass, autopilot, hooks, plugins, MCP, or harness workers only for workspaces whose trust boundary you understand.

Open official documentation

Classification and operating model

Category fit and technical mechanisms are evidence records, not product-quality scores.

Category fit: Qualifies, 4/4 criteria
Operating model: 7/7 layers documented

Inspect category criteria and operating mechanismsFirst-party records

Why it qualifies as a coding harness

This confirms category fit, not product quality. Every required criterion links back to first-party evidence.

Qualifies4 of 4 required criteria evidenced

Adaptive agent loop
Documented
The system repeatedly observes results and chooses the next action instead of following a fixed one-pass graph.
EvidenceAgent loop implementation
Repository tool execution
Documented
The system can use tools to inspect and change a repository or its execution environment.
EvidenceBuilt-in tool registry
Task-aware context management
Documented
The runtime assembles, updates, compacts, retrieves, or persists task-relevant context while work proceeds.
EvidenceContext manager implementation EvidenceProject memory
Model-independent runtime control
Documented
Permissions, budgets, interruption, policy, or stop controls operate outside the model's own text generation.
EvidencePermission modes EvidenceProviders and endpoints

How it works under the hood

Seven mechanisms mapped from first-party records. These labels describe what the harness provides, not how intelligent its model is.

7/7layers documented

Execution & isolationWorkspace isolationDocumented mechanism, not a performance score.
Tooling & integrationsExtensible + browserDocumented mechanism, not a performance score.
Context & statePersistent stateDocumented mechanism, not a performance score.
Lifecycle & recoveryCheckpoint/rewindDocumented mechanism, not a performance score.
ObservabilityStructured tracesDocumented mechanism, not a performance score.
VerificationWorkflow-gatedDocumented mechanism, not a performance score.
Governance & permissionsPolicy controlsDocumented mechanism, not a performance score.

Measured and public context

Configuration-specific measurements and source-native activity stay separate from general product capability.

Inspect code audit, measured configurations, and ecosystem signalsContext, not a product score

Public code audit

3/5public artifacts present

Security policy: Not found
CI workflow: Present at inspected commit
Automated tests: Present at inspected commit
Evaluation assets: Present at inspected commit
Contributor documentation: Not found

The tagged v1.3.187 tree contains 3,878 tracked files, five workflows, 607 test-like paths, and project-owned evaluation scripts. It has no root security policy or contributor guide; internal reviews, engineering tests, and eval utilities do not provide an independent immutable model × harness × environment × budget × attempts result, so no product score is imported.

Inspect commit f99a960149c6d93b8b64e4e08abfc341fbaf7ca2, checked 2026-08-01

Measured configurations

No benchmark run passes the full metadata admission policy for this harness yet. Missing data is not scored as zero.

Benchmark policy and all runs

Context, not quality

Public ecosystem signals

Source-native observations for exact mapped artifacts and reviewed stable release trains. Different units and populations stay separate, and missing coverage is never treated as zero.

View this harness in Usage

Latest stable releasev1.3.187Released 2026-07-31; 213 stable releases in 90 daysOpen release
OpenRouter 30d tokens547.05M#112 coding app; 2026-07-02 to 2026-07-31Open app page
Release asset downloads2.49K299 stable releases; 1794 matched assetsOpen artifact
GitHub stars67Full-source repository; 5 forksOpen artifact

First-party evidence

Each capability claim links to the first-party record that supports it.

18 first-party sourcesProduct record checked 2026-08-01

Product and interfaces

3 sources

GGCode repository overviewOpens in a new tabCoding-agent identity, repository tools, PTC boundary, MCP, multi-agent execution, editor and web surfaces, automation, sessions, and licenseOfficial repositorySource checked 2026-08-01

View 2 more sources

Execution and control

1 source

Permission modesOpens in a new tabSupervised, plan, auto, bypass, and autopilot policies, dangerous-command handling, path rules, approvals, persistence, and goal-directed executionOfficial repositorySource checked 2026-08-01

Agents, state and recovery

2 sources

Project memoryOpens in a new tabAutomatically loaded repository instructions, global and project scopes, persistent save-memory tool, storage, precedence, and cross-session useOfficial repositorySource checked 2026-08-01

View 1 more source

Multi-agent execution modesOpens in a new tabSubagents, persistent swarm teammates, independent harness workers, model selection, tool access, workspace semantics, monitoring, and deliveryOfficial repositorySource checked 2026-08-01

Automation and extensions

4 sources

CLI and daemon modesOpens in a new tabInteractive, pipe, resume, daemon, status, report, ACP, harness, MCP, and programmatic command surfacesOfficial repositorySource checked 2026-08-01

View 3 more sources

Enterprise and operations

1 source

Session execution tracesOpens in a new tabStructured JSON export of LLM calls, tool invocations, timing, token usage, summaries, per-turn events, and offline analysisOfficial repositorySource checked 2026-08-01

Releases and public code audit

7 sources

GGCode v1.3.187 releaseOpens in a new tabLatest stable version, release date, platform packages, checksums, SBOM sidecars, and distribution artifactsOfficial repositorySource checked 2026-08-01

View 6 more sources

Ecosystem context

OpenRouter coding app directory Used to discover GGCode; no OpenRouter rank, traffic value, or category label establishes capability or catalog membership. Observed 2026-08-01.