HarnessMatch
Menu

Coding agent

GitHub Copilot CLI

GitHub-native terminal agent with fleet orchestration and sandbox options.

Status
Active
License
GitHub Copilot CLI License
Evidence
Documented, 16 sources
Product record checked
2026-07-27
Compare harnessesInspect sources

At a glance

GitHub's terminal coding agent supports interactive and programmatic execution, granular permissions, local and cloud sandboxes, MCP, custom agents, parallel subagents, file rollback, and GitHub-hosted or custom local models.

Good choice if

  • GitHub-centered development with repository, issue, pull-request, and MCP context
  • Parallel implementation plans using fleet subagents and separate context windows
  • Teams that need programmatic runs, auditable permissions, rollback, and optional isolation

Check before choosing

  • A Copilot plan or compatible custom provider configuration is required
  • Local and cloud sandbox features are in public preview, opt-in, and have documented platform and lifecycle limits
  • Git-based rewind can restore the entire workspace and delete later files or manual edits; tools-based restoration is experimental and cannot reverse external side effects
See 3 more considerations
  • Copilot Memory is a hosted public-preview feature with a 28-day unused-item retention policy rather than local durable project state
  • Fleet increases GitHub AI Credit use because each subagent calls a model independently
  • The public repository is a support and binary-distribution surface, not the proprietary agent implementation or product evaluation suite

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.

    EvidenceAbout GitHub Copilot CLI

  • Repository tool execution

    Documented

    The system can use tools to inspect and change a repository or its execution environment.

    EvidenceUsing Copilot CLI

  • Task-aware context management

    Documented

    The runtime assembles, updates, compacts, retrieves, or persists task-relevant context while work proceeds.

    EvidenceCopilot Memory

  • Model-independent runtime control

    Documented

    Permissions, budgets, interruption, policy, or stop controls operate outside the model's own text generation.

    EvidenceUsing Copilot CLI

Membership establishes category fit only. It does not score quality, safety, autonomy, model capability, or benchmark performance. · Read the membership rule.

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 & isolationSandbox availableDocumented mechanism, not a performance score.
  • Tooling & integrationsExtensible toolsDocumented mechanism, not a performance score.
  • Context & statePersistent stateDocumented mechanism, not a performance score.
  • Lifecycle & recoveryCheckpoint/rewindDocumented mechanism, not a performance score.
  • ObservabilityLogs/transcriptsDocumented mechanism, not a performance score.
  • VerificationTool-assistedDocumented mechanism, not a performance score.
  • Governance & permissionsPolicy controlsDocumented mechanism, not a performance score.

Public code audit

Unrankedsupport-only repository
Security policy
Not found
CI workflow
Present at inspected commit
Automated tests
Not found
Evaluation assets
Not found
Contributor documentation
Not found

The v1.0.75 public tree contains installers, changelog, issue templates, and ten distribution/release workflows, but no core agent source, product tests, evaluation suite, security policy, or contributor guide. Capability and quality claims therefore cannot be source-audited from this repository.

Inspect commit 05cdbb3ae04a839ffc465facdc0e0f207e37dbe0, checked 2026-07-27

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

Capability support

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

External tools (MCP)
DocumentedProduct-supported MCP integrationSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.
Local models
DocumentedLocal or self-hosted model pathSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.
Agent parallelism
DocumentedDelegated or parallel agent workflowSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.
Runs without an open UI
DocumentedNon-interactive or automation surfaceSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.
Browser control
Not documentedNo first-class support established by the current recordAbsence of current documentation is not proof that the capability is impossible.
Isolated execution
OptionalOptional sandboxed execution modeSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.
Undo file changes
DocumentedProduct-supported file or session rollbackSource · checked 2026-07-27The source establishes the mechanism, not its quality or availability in every mode.

Primary evidence

Each capability claim is tied to a first-party record and a verification date.

Product record checked 2026-07-27
About GitHub Copilot CLITerminal workflow, Copilot plans, MCP, custom agents, memory, custom providers, and local modelsOfficial docsProgrammatic referenceHeadless execution, model pinning, tool and URL permissions, secrets redaction, and transcriptsOfficial docsCloud and local sandboxesFilesystem, network, and system isolation plus preview status and platform limitationsOfficial docsFleet subagentsTask decomposition, parallel subagents, model selection, context isolation, and orchestrationOfficial docsSession rollbackConversation rewind, Git snapshots, tool-based file restore, constraints, and destructive effectsOfficial docsCopilot MemoryPublic-preview repository facts, user preferences, CLI use, citations, validation, enablement, retention, and governanceOfficial docsUsing Copilot CLIInteractive modes, approval defaults, tool execution, sessions, autonomy, and day-to-day workflowOfficial docsLocal and cloud sandboxesPreview status, local OS isolation, hosted cloud environments, boundaries, billing, and platform constraintsOfficial docs
View 8 additional sources
Additional first-party evidence8 sources
Custom providersOpenAI-compatible, Anthropic, Azure, and local Ollama provider configuration and limitationsOfficial docsMCP configurationBuilt-in GitHub MCP, additional servers, transports, tools, authentication, trust, and managementOfficial docsAutopilot modeAutonomous continuation, programmatic use, permission interaction, explicit continuation limits, and stop conditionsOfficial docsCopilot hooksPre-tool policy decisions, lifecycle events, subagent completion, audit logging, validation, and security cautionsOfficial docsCLI customization overviewCustom instructions, hooks, skills, custom subagents, MCP, plugins, and personal settingsOfficial docsCustom agents and subagentsAgent profiles, scoped tools and MCP, built-in agents, separate contexts, delegation, and reusable specializationOfficial docsHooks referenceLifecycle event schemas, pre-tool and permission decisions, stop controls, matcher filtering, outputs, and failure behaviorOfficial docsCopilot CLI v1.0.75 support repositoryPinned installer and changelog support tree, release workflows, distribution license, and absent core sourceOfficial repository