GitHub collection

arbiterforge/codearbiter

Imported from GitHub — is this yours?

163 skills imported from this repository.

View on GitHub
145 7 AGPL-3.0
Claude Skill

brainstorming

The Socratic spec-refinement front of /feature, and the planning front of /sprint. Routed to BEFORE any code — it takes a one-line idea and drives it to an approved, concrete spec with testable acceptance criteria. Five gated phases — frame, shape, refine, write, review-and-appro

LLM Mart

Claude Skill

commit-gate

The only path to a commit. Routed to when the user invokes /commit or otherwise instructs codeArbiter to persist staged changes. Nine gated phases — permission, branch, classification, verification (test/lint/secrets), behavioral proof, diff review, selective stage, message, comm

LLM Mart

Claude Skill

context-check

Optional manual drift audit — report stale provenance-tracked docs (via _provenancelib drift detection across .codearbiter/.provenance/), then per stale doc offer re-scout / re-baseline / defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill

context-creation

The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase t

LLM Mart

Claude Skill

crypto-compliance

The banned-primitive gate. Routed to when changed code hashes, signs, encrypts, derives keys, generates security-relevant randomness, configures TLS, or imports a crypto library. Rejects broken primitives, disabled TLS verification, and home-rolled crypto; the approved-primitive

LLM Mart

Claude Skill

debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.

LLM Mart

Claude Skill

decision-lifecycle

Author and track Architecture Decision Records. Routed to when the user invokes /adr to record a new decision or /adr-status to list ADR health. Authors numbered, dated, user-attributed ADRs under .codearbiter/decisions/, maintains supersede chains, and reports status read-only.

LLM Mart

Claude Skill

decision-variance

Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report

LLM Mart

Claude Skill

decompose

The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk

LLM Mart

Claude Skill

dispatching-parallel-agents

The parallel fan-out primitive. Routed to by any skill or command that splits work across independent units and dispatches an agent per unit — subagent-driven-development, /sprint, parallel /review. It owns the dispatch/collect/funnel discipline: bound concurrency, isolate units,

LLM Mart

Claude Skill

executing-plans

The checkpoint coordinator for /feature. Routed to by /feature once a writing-plans plan exists. Groups tasks into batches, delegates each batch to subagent-driven-development (fresh author agent per task, full review chain, fresh verification), then stops for a human checkpoint

LLM Mart

Claude Skill

finishing-a-development-branch

The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" a

LLM Mart

Claude Skill

post-merge-cleanup

Finish an already-merged branch. Proves the branch is contained in the fetched default, classifies leftover artifacts as unique / redundant / superseded, returns to a clean --ff-only default checkout, and deletes the merged local branch — every discard confirmed per item. Routed

LLM Mart

Claude Skill

refactor

The behavior-preserving restructure gate. Routed to by /refactor for a rename, extract, inline, move, dedup, or internal-implementation swap. Six gated phases prove behavioral parity through unmodified pre-existing tests; any diff that classifies as `feat` is not a refactor and i

LLM Mart

Claude Skill

release

The single permitted path to a version tag, for any target declared in the project's .codearbiter/release-targets.md. Routed to when the user invokes /release on a non-default branch with a green suite. Takes the declared target as its one argument, derives the SemVer bump from C

LLM Mart

Claude Skill

secret-handling

The secret-source gate. Routed to when changed code reads, writes, or passes a secret — API key, token, password, connection string, signing key, certificate, or any value that grants access. Validates that every secret comes from the approved store and never lands in source, log

LLM Mart

Claude Skill

security-architecture

Optional, opt-in STRIDE threat pass for a sensitive feature — invoked deliberately via /threat-model, never forced on ordinary changes. Walks the change's attack surface and security boundaries (governed by {{PROJECT_DIR}}/.codearbiter/security-controls.md), surfaces threats and

LLM Mart

Claude Skill

skill-author

The authoring gate for new skills. Routed to when the user invokes /new-skill "<gap>". Five gated phases — gap evidence, scope, authoring, self-review against the v2 house style, routing integration. A new skill is not written until an existing one is proven not to cover the gap,

LLM Mart

Claude Skill

tdd

The test-first gate. Routed to by /feature (after the spec is approved), /fix, and /refactor before any implementation code is written. Six gated phases — obligation scan, red, green, obligation verify, coverage, lint. No feature code exists before Phase 1 clears; nothing reaches

LLM Mart

Claude Skill

tribunal

The deep, rarely-convened whole-codebase audit lane. Routed to when the user invokes {{CMD:tribunal}}. Seven gated phases — cost/model, map, roster dispatch, triage, report, approval+filing, telemetry. Costs on the order of millions of tokens; proceeds only after the user acknowl

LLM Mart

Claude Skill

using-git-worktrees

OPTIONAL per-task isolation for autonomous parallel work. Routed to only on explicit opt-in by subagent-driven-development or dispatching-parallel-agents, so parallel units mutate files without colliding. Stands up one worktree per unit, works it in isolation, then integrates eac

LLM Mart

Claude Skill

writing-plans

The spec-to-plan bridge. Routed to by /feature once the brainstormed spec is approved, and by /sprint before execution. Decomposes the spec into 2–5 minute tasks, each carrying its exact file path(s) and a concrete verification step that maps to a tdd obligation. Writes the plan

LLM Mart

Claude Skill

brainstorming

The Socratic spec-refinement front of /feature, and the planning front of /sprint. Routed to BEFORE any code — it takes a one-line idea and drives it to an approved, concrete spec with testable acceptance criteria. Five gated phases — frame, shape, refine, write, review-and-appro

LLM Mart

Claude Skill

commit-gate

The only path to a commit. Routed to when the user invokes /commit or otherwise instructs codeArbiter to persist staged changes. Nine gated phases — permission, branch, classification, verification (test/lint/secrets), behavioral proof, diff review, selective stage, message, comm

LLM Mart

Claude Skill

context-check

Optional manual drift audit — report stale provenance-tracked docs (via _provenancelib drift detection across .codearbiter/.provenance/), then per stale doc offer re-scout / re-baseline / defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill

context-creation

The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase t

LLM Mart

Claude Skill

crypto-compliance

The banned-primitive gate. Routed to when changed code hashes, signs, encrypts, derives keys, generates security-relevant randomness, configures TLS, or imports a crypto library. Rejects broken primitives, disabled TLS verification, and home-rolled crypto; the approved-primitive

LLM Mart

Claude Skill

debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.

LLM Mart

Claude Skill

decision-lifecycle

Author and track Architecture Decision Records. Routed to when the user invokes /adr to record a new decision or /adr-status to list ADR health. Authors numbered, dated, user-attributed ADRs under .codearbiter/decisions/, maintains supersede chains, and reports status read-only.

LLM Mart

Claude Skill

decision-variance

Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report

LLM Mart

Claude Skill

decompose

The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk

LLM Mart

Claude Skill

dispatching-parallel-agents

The parallel fan-out primitive. Routed to by any skill or command that splits work across independent units and dispatches an agent per unit — subagent-driven-development, /sprint, parallel /review. It owns the dispatch/collect/funnel discipline: bound concurrency, isolate units,

LLM Mart

Claude Skill

executing-plans

The checkpoint coordinator for /feature. Routed to by /feature once a writing-plans plan exists. Groups tasks into batches, delegates each batch to subagent-driven-development (fresh author agent per task, full review chain, fresh verification), then stops for a human checkpoint

LLM Mart

Claude Skill

finishing-a-development-branch

The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" a

LLM Mart

Claude Skill

post-merge-cleanup

Finish an already-merged branch. Proves the branch is contained in the fetched default, classifies leftover artifacts as unique / redundant / superseded, returns to a clean --ff-only default checkout, and deletes the merged local branch — every discard confirmed per item. Routed

LLM Mart

Claude Skill

refactor

The behavior-preserving restructure gate. Routed to by /refactor for a rename, extract, inline, move, dedup, or internal-implementation swap. Six gated phases prove behavioral parity through unmodified pre-existing tests; any diff that classifies as `feat` is not a refactor and i

LLM Mart

Claude Skill

release

The single permitted path to a version tag, for any target declared in the project's .codearbiter/release-targets.md. Routed to when the user invokes /release on a non-default branch with a green suite. Takes the declared target as its one argument, derives the SemVer bump from C

LLM Mart

Claude Skill

secret-handling

The secret-source gate. Routed to when changed code reads, writes, or passes a secret — API key, token, password, connection string, signing key, certificate, or any value that grants access. Validates that every secret comes from the approved store and never lands in source, log

LLM Mart

Claude Skill

security-architecture

Optional, opt-in STRIDE threat pass for a sensitive feature — invoked deliberately via /threat-model, never forced on ordinary changes. Walks the change's attack surface and security boundaries (governed by <project-root>/.codearbiter/security-controls.md), surfaces threats and u

LLM Mart

Claude Skill

skill-author

The authoring gate for new skills. Routed to when the user invokes /new-skill "<gap>". Five gated phases — gap evidence, scope, authoring, self-review against the v2 house style, routing integration. A new skill is not written until an existing one is proven not to cover the gap,

LLM Mart

Claude Skill

tdd

The test-first gate. Routed to by /feature (after the spec is approved), /fix, and /refactor before any implementation code is written. Six gated phases — obligation scan, red, green, obligation verify, coverage, lint. No feature code exists before Phase 1 clears; nothing reaches

LLM Mart

Claude Skill

tribunal

The deep, rarely-convened whole-codebase audit lane. Routed to when the user invokes $ca-tribunal. Seven gated phases — cost/model, map, roster dispatch, triage, report, approval+filing, telemetry. Costs on the order of millions of tokens; proceeds only after the user acknowledge

LLM Mart

Claude Skill

using-git-worktrees

OPTIONAL per-task isolation for autonomous parallel work. Routed to only on explicit opt-in by subagent-driven-development or dispatching-parallel-agents, so parallel units mutate files without colliding. Stands up one worktree per unit, works it in isolation, then integrates eac

LLM Mart

Claude Skill

writing-plans

The spec-to-plan bridge. Routed to by /feature once the brainstormed spec is approved, and by /sprint before execution. Decomposes the spec into 2–5 minute tasks, each carrying its exact file path(s) and a concrete verification step that maps to a tdd obligation. Writes the plan

LLM Mart

Claude Skill

ca-add-dep

Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs.

LLM Mart

Claude Skill

ca-adr

Author a numbered, dated, user-attributed Architecture Decision Record under .codearbiter/decisions/.

LLM Mart

Claude Skill

ca-adr-status

Report the health of Architecture Decision Records — aged, unchallenged, supersession candidates, unresolved CONFIRM-NN. Read-only.

LLM Mart

Claude Skill

ca-audit

Assemble the governance record for a range — commits, overrides, ADRs, sprint auto-decisions, open questions, checkpoint findings — into one dated audit packet. Read-only.

LLM Mart

Claude Skill

ca-btw

Lightweight Q&A about the project — answer from context and return, no routing, no state change.

LLM Mart

Claude Skill

ca-checkpoint

Periodic multi-reviewer sweep of the whole codebase — surfaces a triaged checkpoint report.

LLM Mart

Claude Skill

ca-chore

Sanctioned lane for non-behavioral work — docs-only edits, dependency bumps, reverts. Type-scaled gates; no TDD demanded of prose.

LLM Mart

Claude Skill

ca-cleanup

Finish an already-merged branch — classify the leftover artifacts, return to a fast-forwarded default checkout, and delete the merged local branch. Every discard confirmed per item; containment proven, never assumed.

LLM Mart

Claude Skill

ca-commands

Show the codeArbiter command catalog — the public command list and what each routes to.

LLM Mart

Claude Skill

ca-commit

Run the full commit gate — the only sanctioned path to a git commit.

LLM Mart

Claude Skill

ca-conflict

Stop everything and surface a rule conflict — persona vs. docs vs. code. Present both sides and the conflict-hierarchy level; the user resolves. No silent reconciliation.

LLM Mart

Claude Skill

ca-context-check

Optional manual drift audit — report stale provenance-tracked docs, then per stale doc offer re-scout, re-baseline, or defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill

ca-create-context

Brownfield back-fill — scout an existing codebase and populate .codearbiter/, then lock it initialized.

LLM Mart

Claude Skill

ca-debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown. No code changes — exits to $ca-fix, $ca-adr, or a no-action close.

LLM Mart

Claude Skill

ca-decompose

Greenfield decomposition interview — a layered interview that populates .codearbiter/ and locks it initialized.

LLM Mart

Claude Skill

ca-doctor

Verify the active host install, package, command ownership, enforcement, and harmless live-fire probe. Read-only.

LLM Mart

Claude Skill

ca-feature

Start a feature: brainstorm a spec, get it approved, then drive it test-first through the pipeline. The one entry to implementation.

LLM Mart

Claude Skill

ca-fix

Fix a confirmed bug: a failing regression test first, then a minimal fix, then the rest of the tdd gates.

LLM Mart

Claude Skill

ca-init

Opt this repo into codeArbiter — scaffold the root-level .codearbiter/ state store.

LLM Mart

Claude Skill

ca-metrics

Read-only 3-metric governance glance — override rate, small-lane rate, sprint low-confidence ratio — each with a trend arrow vs. the prior 20-commit window.

LLM Mart

Claude Skill

ca-new-skill

Author a new codeArbiter skill: prove the gap is real, get the spec approved, then write it.

LLM Mart

Claude Skill

ca-override

Sanctioned, logged bypass of a gate or hard rule — one audit line, then proceed.

LLM Mart

Claude Skill

ca-pr

Open a pull request the only sanctioned way — clear every BLOCK-level review finding, then stage the PR. Never a direct write to the default branch.

LLM Mart

Claude Skill

ca-preview

Zero-onboarding, read-only dry-run of the reviewer fleet against the current uncommitted diff. Predicts reviewers, runs the state-free secret scan, writes nothing.

LLM Mart

Claude Skill

ca-reconcile

SMARTS arbitration — reconcile architectural artifacts against the scaffold and prior decisions; every variance resolved by an explicit, user-attributed choice.

LLM Mart

Claude Skill

ca-refactor

Restructure code with behavioral parity proven through unmodified pre-existing tests, then refactor. No behavior change.

LLM Mart

Claude Skill

ca-release

Cut a release the only sanctioned way — SemVer bump from the commit log, a CHANGELOG section, an annotated tag. Takes the declared target's name as its only argument, or --dry-run to preview one with no write. The only path to a version tag.

LLM Mart

Claude Skill

ca-review

Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR.

LLM Mart

Claude Skill

ca-spike

Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or $ca-feature.

LLM Mart

Claude Skill

ca-sprint

Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops.

LLM Mart

Claude Skill

ca-standup

Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes.

LLM Mart

Claude Skill

ca-status

Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only.

LLM Mart

Claude Skill

ca-task

The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md.

LLM Mart

Claude Skill

ca-threat-model

Opt-in lightweight STRIDE pass for a sensitive feature before implementation. Not a routine gate — invoke it when a change warrants security thought.

LLM Mart

Claude Skill

ca-tribunal

Deep, rarely-convened whole-codebase audit — eleven specialist lenses, a resumable on-disk audit log, findings filed as GitHub issues on approval. Expensive; estimates cost and STOPs before running. Never a required gate.

LLM Mart

Claude Skill

ca-watch

Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges.

LLM Mart

Claude Skill

brainstorming

The Socratic spec-refinement front of /feature, and the planning front of /sprint. Routed to BEFORE any code — it takes a one-line idea and drives it to an approved, concrete spec with testable acceptance criteria. Five gated phases — frame, shape, refine, write, review-and-appro

LLM Mart

Claude Skill

commit-gate

The only path to a commit. Routed to when the user invokes /commit or otherwise instructs codeArbiter to persist staged changes. Nine gated phases — permission, branch, classification, verification (test/lint/secrets), behavioral proof, diff review, selective stage, message, comm

LLM Mart

Claude Skill

context-check

Optional manual drift audit — report stale provenance-tracked docs (via _provenancelib drift detection across .codearbiter/.provenance/), then per stale doc offer re-scout / re-baseline / defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill

context-creation

The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase t

LLM Mart

Claude Skill

crypto-compliance

The banned-primitive gate. Routed to when changed code hashes, signs, encrypts, derives keys, generates security-relevant randomness, configures TLS, or imports a crypto library. Rejects broken primitives, disabled TLS verification, and home-rolled crypto; the approved-primitive

LLM Mart

Claude Skill

debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.

LLM Mart

Claude Skill

decision-lifecycle

Author and track Architecture Decision Records. Routed to when the user invokes /adr to record a new decision or /adr-status to list ADR health. Authors numbered, dated, user-attributed ADRs under .codearbiter/decisions/, maintains supersede chains, and reports status read-only.

LLM Mart

Claude Skill

decision-variance

Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report

LLM Mart

Claude Skill

decompose

The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk

LLM Mart

Claude Skill

dispatching-parallel-agents

The parallel fan-out primitive. Routed to by any skill or command that splits work across independent units and dispatches an agent per unit — subagent-driven-development, /sprint, parallel /review. It owns the dispatch/collect/funnel discipline: bound concurrency, isolate units,

LLM Mart

Claude Skill

executing-plans

The checkpoint coordinator for /feature. Routed to by /feature once a writing-plans plan exists. Groups tasks into batches, delegates each batch to subagent-driven-development (fresh author agent per task, full review chain, fresh verification), then stops for a human checkpoint

LLM Mart

Claude Skill

finishing-a-development-branch

The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" a

LLM Mart

Claude Skill

post-merge-cleanup

Finish an already-merged branch. Proves the branch is contained in the fetched default, classifies leftover artifacts as unique / redundant / superseded, returns to a clean --ff-only default checkout, and deletes the merged local branch — every discard confirmed per item. Routed

LLM Mart

Claude Skill

refactor

The behavior-preserving restructure gate. Routed to by /refactor for a rename, extract, inline, move, dedup, or internal-implementation swap. Six gated phases prove behavioral parity through unmodified pre-existing tests; any diff that classifies as `feat` is not a refactor and i

LLM Mart

Claude Skill

release

The single permitted path to a version tag, for any target declared in the project's .codearbiter/release-targets.md. Routed to when the user invokes /release on a non-default branch with a green suite. Takes the declared target as its one argument, derives the SemVer bump from C

LLM Mart

Claude Skill

secret-handling

The secret-source gate. Routed to when changed code reads, writes, or passes a secret — API key, token, password, connection string, signing key, certificate, or any value that grants access. Validates that every secret comes from the approved store and never lands in source, log

LLM Mart

Claude Skill

security-architecture

Optional, opt-in STRIDE threat pass for a sensitive feature — invoked deliberately via /threat-model, never forced on ordinary changes. Walks the change's attack surface and security boundaries (governed by <project-root>/.codearbiter/security-controls.md), surfaces threats and u

LLM Mart

Claude Skill

skill-author

The authoring gate for new skills. Routed to when the user invokes /new-skill "<gap>". Five gated phases — gap evidence, scope, authoring, self-review against the v2 house style, routing integration. A new skill is not written until an existing one is proven not to cover the gap,

LLM Mart

Claude Skill

tdd

The test-first gate. Routed to by /feature (after the spec is approved), /fix, and /refactor before any implementation code is written. Six gated phases — obligation scan, red, green, obligation verify, coverage, lint. No feature code exists before Phase 1 clears; nothing reaches

LLM Mart

Claude Skill

tribunal

The deep, rarely-convened whole-codebase audit lane. Routed to when the user invokes /ca-tribunal. Seven gated phases — cost/model, map, roster dispatch, triage, report, approval+filing, telemetry. Costs on the order of millions of tokens; proceeds only after the user acknowledge

LLM Mart

Claude Skill

using-git-worktrees

OPTIONAL per-task isolation for autonomous parallel work. Routed to only on explicit opt-in by subagent-driven-development or dispatching-parallel-agents, so parallel units mutate files without colliding. Stands up one worktree per unit, works it in isolation, then integrates eac

LLM Mart

Claude Skill

writing-plans

The spec-to-plan bridge. Routed to by /feature once the brainstormed spec is approved, and by /sprint before execution. Decomposes the spec into 2–5 minute tasks, each carrying its exact file path(s) and a concrete verification step that maps to a tdd obligation. Writes the plan

LLM Mart

Claude Skill

ca-add-dep

Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs.

LLM Mart

Claude Skill

ca-adr

Author a numbered, dated, user-attributed Architecture Decision Record under .codearbiter/decisions/.

LLM Mart

Claude Skill

ca-adr-status

Report the health of Architecture Decision Records — aged, unchallenged, supersession candidates, unresolved CONFIRM-NN. Read-only.

LLM Mart

Claude Skill

ca-audit

Assemble the governance record for a range — commits, overrides, ADRs, sprint auto-decisions, open questions, checkpoint findings — into one dated audit packet. Read-only.

LLM Mart

Claude Skill

ca-btw

Lightweight Q&A about the project — answer from context and return, no routing, no state change.

LLM Mart

Claude Skill

ca-checkpoint

Periodic multi-reviewer sweep of the whole codebase — surfaces a triaged checkpoint report.

LLM Mart

Claude Skill

ca-chore

Sanctioned lane for non-behavioral work — docs-only edits, dependency bumps, reverts. Type-scaled gates; no TDD demanded of prose.

LLM Mart

Claude Skill

ca-cleanup

Finish an already-merged branch — classify the leftover artifacts, return to a fast-forwarded default checkout, and delete the merged local branch. Every discard confirmed per item; containment proven, never assumed.

LLM Mart

Claude Skill

ca-commands

Show the codeArbiter command catalog — the public command list and what each routes to.

LLM Mart

Claude Skill

ca-commit

Run the full commit gate — the only sanctioned path to a git commit.

LLM Mart

Claude Skill

ca-conflict

Stop everything and surface a rule conflict — persona vs. docs vs. code. Present both sides and the conflict-hierarchy level; the user resolves. No silent reconciliation.

LLM Mart

Claude Skill

ca-context-check

Optional manual drift audit — report stale provenance-tracked docs, then per stale doc offer re-scout, re-baseline, or defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill

ca-create-context

Brownfield back-fill — scout an existing codebase and populate .codearbiter/, then lock it initialized.

LLM Mart

Claude Skill

ca-debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown. No code changes — exits to /ca-fix, /ca-adr, or a no-action close.

LLM Mart

Claude Skill

ca-decompose

Greenfield decomposition interview — a layered interview that populates .codearbiter/ and locks it initialized.

LLM Mart

Claude Skill

ca-doctor

Verify the active host install, package, command ownership, enforcement, wrapper self-test, and active-dispatch coverage gap. Read-only.

LLM Mart

Claude Skill

ca-feature

Start a feature: brainstorm a spec, get it approved, then drive it test-first through the pipeline. The one entry to implementation.

LLM Mart

Claude Skill

ca-fix

Fix a confirmed bug: a failing regression test first, then a minimal fix, then the rest of the tdd gates.

LLM Mart

Claude Skill

ca-init

Opt this repo into codeArbiter — scaffold the root-level .codearbiter/ state store.

LLM Mart

Claude Skill

ca-metrics

Read-only 3-metric governance glance — override rate, small-lane rate, sprint low-confidence ratio — each with a trend arrow vs. the prior 20-commit window.

LLM Mart

Claude Skill

ca-new-skill

Author a new codeArbiter skill: prove the gap is real, get the spec approved, then write it.

LLM Mart

Claude Skill

ca-override

Sanctioned, logged bypass of a gate or hard rule — one audit line, then proceed.

LLM Mart

Claude Skill

ca-pr

Open a pull request the only sanctioned way — clear every BLOCK-level review finding, then stage the PR. Never a direct write to the default branch.

LLM Mart

Claude Skill

ca-preview

Zero-onboarding, read-only dry-run of the reviewer fleet against the current uncommitted diff. Predicts reviewers, runs the state-free secret scan, writes nothing.

LLM Mart

Claude Skill

ca-prune

Trim transcript clutter to extend session lifetime — analyze, prune a copy, or toggle the after-each-turn service. Dry-run by default; gains land at resume/compaction, not the current turn.

LLM Mart

Claude Skill

ca-reconcile

SMARTS arbitration — reconcile architectural artifacts against the scaffold and prior decisions; every variance resolved by an explicit, user-attributed choice.

LLM Mart

Claude Skill

ca-refactor

Restructure code with behavioral parity proven through unmodified pre-existing tests, then refactor. No behavior change.

LLM Mart

Claude Skill

ca-release

Cut a release the only sanctioned way — SemVer bump from the commit log, a CHANGELOG section, an annotated tag. Takes the declared target's name as its only argument, or --dry-run to preview one with no write. The only path to a version tag.

LLM Mart

Claude Skill

ca-review

Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR.

LLM Mart

Claude Skill

ca-spike

Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or /ca-feature.

LLM Mart

Claude Skill

ca-sprint

Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops.

LLM Mart

Claude Skill

ca-standup

Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes.

LLM Mart

Claude Skill

ca-status

Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only.

LLM Mart

Claude Skill

ca-task

The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md.

LLM Mart

Claude Skill

ca-threat-model

Opt-in lightweight STRIDE pass for a sensitive feature before implementation. Not a routine gate — invoke it when a change warrants security thought.

LLM Mart

Claude Skill

ca-tribunal

Deep, rarely-convened whole-codebase audit — eleven specialist lenses, a resumable on-disk audit log, findings filed as GitHub issues on approval. Expensive; estimates cost and STOPs before running. Never a required gate.

LLM Mart

Claude Skill

ca-watch

Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges.

LLM Mart

Claude Skill ca

brainstorming

The Socratic spec-refinement front of /feature, and the planning front of /sprint. Routed to BEFORE any code — it takes a one-line idea and drives it to an approved, concrete spec with testable acceptance criteria. Five gated phases — frame, shape, refine, write, review-and-appro

LLM Mart

Claude Skill ca

commit-gate

The only path to a commit. Routed to when the user invokes /commit or otherwise instructs codeArbiter to persist staged changes. Nine gated phases — permission, branch, classification, verification (test/lint/secrets), behavioral proof, diff review, selective stage, message, comm

LLM Mart

Claude Skill ca

context-check

Optional manual drift audit — report stale provenance-tracked docs (via _provenancelib drift detection across .codearbiter/.provenance/), then per stale doc offer re-scout / re-baseline / defer. Not the daily loop; commit-gate auto-heal owns routine maintenance.

LLM Mart

Claude Skill ca

context-creation

The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase t

LLM Mart

Claude Skill ca

crypto-compliance

The banned-primitive gate. Routed to when changed code hashes, signs, encrypts, derives keys, generates security-relevant randomness, configures TLS, or imports a crypto library. Rejects broken primitives, disabled TLS verification, and home-rolled crypto; the approved-primitive

LLM Mart

Claude Skill ca

debug

Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.

LLM Mart

Claude Skill ca

decision-lifecycle

Author and track Architecture Decision Records. Routed to when the user invokes /adr to record a new decision or /adr-status to list ADR health. Authors numbered, dated, user-attributed ADRs under .codearbiter/decisions/, maintains supersede chains, and reports status read-only.

LLM Mart

Claude Skill ca

decision-variance

Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report

LLM Mart

Claude Skill ca

decompose

The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk

LLM Mart

Claude Skill ca

dispatching-parallel-agents

The parallel fan-out primitive. Routed to by any skill or command that splits work across independent units and dispatches an agent per unit — subagent-driven-development, /sprint, parallel /review. It owns the dispatch/collect/funnel discipline: bound concurrency, isolate units,

LLM Mart

Claude Skill ca

executing-plans

The checkpoint coordinator for /feature. Routed to by /feature once a writing-plans plan exists. Groups tasks into batches, delegates each batch to subagent-driven-development (fresh author agent per task, full review chain, fresh verification), then stops for a human checkpoint

LLM Mart

Claude Skill ca

finishing-a-development-branch

The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" a

LLM Mart

Claude Skill ca

post-merge-cleanup

Finish an already-merged branch. Proves the branch is contained in the fetched default, classifies leftover artifacts as unique / redundant / superseded, returns to a clean --ff-only default checkout, and deletes the merged local branch — every discard confirmed per item. Routed

LLM Mart

Claude Skill ca

refactor

The behavior-preserving restructure gate. Routed to by /refactor for a rename, extract, inline, move, dedup, or internal-implementation swap. Six gated phases prove behavioral parity through unmodified pre-existing tests; any diff that classifies as `feat` is not a refactor and i

LLM Mart

Claude Skill ca

release

The single permitted path to a version tag, for any target declared in the project's .codearbiter/release-targets.md. Routed to when the user invokes /release on a non-default branch with a green suite. Takes the declared target as its one argument, derives the SemVer bump from C

LLM Mart

Claude Skill ca

secret-handling

The secret-source gate. Routed to when changed code reads, writes, or passes a secret — API key, token, password, connection string, signing key, certificate, or any value that grants access. Validates that every secret comes from the approved store and never lands in source, log

LLM Mart

Claude Skill ca

security-architecture

Optional, opt-in STRIDE threat pass for a sensitive feature — invoked deliberately via /threat-model, never forced on ordinary changes. Walks the change's attack surface and security boundaries (governed by ${CLAUDE_PROJECT_DIR}/.codearbiter/security-controls.md), surfaces threat

LLM Mart

Claude Skill ca

skill-author

The authoring gate for new skills. Routed to when the user invokes /new-skill "<gap>". Five gated phases — gap evidence, scope, authoring, self-review against the v2 house style, routing integration. A new skill is not written until an existing one is proven not to cover the gap,

LLM Mart

Claude Skill ca

tdd

The test-first gate. Routed to by /feature (after the spec is approved), /fix, and /refactor before any implementation code is written. Six gated phases — obligation scan, red, green, obligation verify, coverage, lint. No feature code exists before Phase 1 clears; nothing reaches

LLM Mart

Claude Skill ca

tribunal

The deep, rarely-convened whole-codebase audit lane. Routed to when the user invokes /ca:tribunal. Seven gated phases — cost/model, map, roster dispatch, triage, report, approval+filing, telemetry. Costs on the order of millions of tokens; proceeds only after the user acknowledge

LLM Mart

Claude Skill ca

using-git-worktrees

OPTIONAL per-task isolation for autonomous parallel work. Routed to only on explicit opt-in by subagent-driven-development or dispatching-parallel-agents, so parallel units mutate files without colliding. Stands up one worktree per unit, works it in isolation, then integrates eac

LLM Mart

Claude Skill ca

writing-plans

The spec-to-plan bridge. Routed to by /feature once the brainstormed spec is approved, and by /sprint before execution. Decomposes the spec into 2–5 minute tasks, each carrying its exact file path(s) and a concrete verification step that maps to a tdd obligation. Writes the plan

LLM Mart

Claude Skill ca-sandbox

sandbox-claude-inside

Run Claude Code INSIDE a ca-sandbox box (`--with-claude`). Routed to when the user wants an agent loop running against an isolated, ephemeral sandbox rather than the host. Authenticates via an env-injected CLAUDE_CODE_OAUTH_TOKEN with no host bind of ~/.claude; the image pins the

LLM Mart

Claude Skill ca-sandbox

sandbox-lifecycle

The lifecycle gate for a local Codespace-equivalent sandbox. Routed to when the user invokes /ca-sandbox:sandbox to pull an untrusted repo into an ephemeral, host-FS-isolated Docker container, or any of the interaction commands (/ca-sandbox:sandbox-shell, /ca-sandbox:sandbox-exec

LLM Mart

Agents (75)

Claude Agent

architecture-drift-reviewer

Read-only checkpoint reviewer. Surfaces drift between the codebase and accepted ADRs in .codearbiter/decisions/. Informational — never blocks.

Claude Agent

auth-crypto-reviewer

Reviews authentication, cryptography, key handling, and secrets against {{PROJECT_DIR}}/.codearbiter/security-controls.md. Hard blocks on banned primitives, exposed secrets, disabled TLS verification, and shell injection. Read-only checkpoint reviewer.

Claude Agent

backend-author

Use when writing or modifying backend/server-side code. Owns the TDD workflow, input validation, framework conventions, and ORM usage. MUST write failing tests before implementation code. Reads tech stack from {{PROJECT_DIR}}/.codearbiter/tech-stack.md.

Claude Agent

checkpoint-aggregator

Composes the finding-triage report and decision-challenger output into a dated checkpoint document under .codearbiter/checkpoints/YYYY-MM-DD.md. Aggregator, not a blocker.

Claude Agent

coverage-auditor

Dispatched by the tdd skill (Phase 4) to audit test coverage against TDD obligations. Identifies untested source files, coverage below the maturity threshold, and logical test gaps.

Claude Agent

decision-challenger

Adversarial red-team reviewer of ADRs. Builds the strongest case against each decision, names load-bearing assumptions, assigns confidence 1–5, and surfaces evidence that would prove a decision wrong. Read-only. Dispatched optionally by decision-variance. Reads ADRs from .codearb

Claude Agent

dependency-reviewer

Dispatched when package.json, lock files, or container base images change. Verifies license, provenance, maintenance signal, and supply-chain posture against .codearbiter/security-controls.md and .codearbiter/tech-stack.md before merge.

Claude Agent

design-quality-reviewer

Reviews generated, user-facing visual or formatted output (UI, reports, slides, charts, diagrams, CLI output) against the anti-slop-design reference. Read-only; surfaces design-slop findings. Dispatched by frontend-author on UI changes. Tier 2 document producers (/pr, release) ap

Claude Agent

finding-triage

Post-processes all checkpoint reviewer reports — consolidates findings, classifies each by severity and whether it blocks the current change. Sequential. Produces a unified triage report for the checkpoint-aggregator. Reads project state from .codearbiter/.

Claude Agent

frontend-author

Use when writing or modifying frontend/UI code. Owns the TDD workflow, component conventions, state management, and UI security. MUST write failing tests before implementation code. Reads tech stack from {{PROJECT_DIR}}/.codearbiter/tech-stack.md.

Claude Agent

grader

INTERNAL SMARTS analyst dispatched by the decision-variance skill. Produces a SMARTS analysis and recommendation for one (artifact-position, scaffold-evidence) pair. Never decides — the user decides. Never dispatch directly.

Claude Agent

INDEX

Imported from arbiterforge/codearbiter/core/surface/agents/INDEX.md.

Claude Agent

infra-author

Use when writing or modifying IaC, containers, CI/CD manifests, or deployment configuration. Reads tech stack and security boundaries from {{PROJECT_DIR}}/.codearbiter/.

Claude Agent

map-deps

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload dependency/integration-surface mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

map-structure

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload structural mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

migration-reviewer

Dispatched when a database migration file is added or modified. Reviews migrations for safety, data-classification tagging, and immutability against .codearbiter/security-controls.md.

Claude Agent

scout

INTERNAL evidence-gatherer dispatched by the decision-variance and context-creation skills. Scans an assigned code scope and reports evidence of architectural decisions — file paths and line numbers only. Never dispatch directly.

Claude Agent

security-reviewer

Dispatch PROACTIVELY when a change touches authentication, authorization, cryptography, secrets, deployment manifests, network policies, or CI workflows. Reviews diffs against {{PROJECT_DIR}}/.codearbiter/security-controls.md. Read-only; produces findings.

Claude Agent

tribunal-lens-reviewer

Dispatched by the tribunal deep-audit lane, once per active lens, with a lens assignment. Generic read-only lens executor — reads the assigned lens mandate under skills/tribunal/references/lenses/ and writes one file per finding. Never dispatch directly.

Claude Agent

architecture-drift-reviewer

Read-only checkpoint reviewer. Surfaces drift between the codebase and accepted ADRs in .codearbiter/decisions/. Informational — never blocks.

Claude Agent

auth-crypto-reviewer

Reviews authentication, cryptography, key handling, and secrets against ${CLAUDE_PROJECT_DIR}/.codearbiter/security-controls.md. Hard blocks on banned primitives, exposed secrets, disabled TLS verification, and shell injection. Read-only checkpoint reviewer.

Claude Agent

backend-author

Use when writing or modifying backend/server-side code. Owns the TDD workflow, input validation, framework conventions, and ORM usage. MUST write failing tests before implementation code. Reads tech stack from ${CLAUDE_PROJECT_DIR}/.codearbiter/tech-stack.md.

Claude Agent

checkpoint-aggregator

Composes the finding-triage report and decision-challenger output into a dated checkpoint document under .codearbiter/checkpoints/YYYY-MM-DD.md. Aggregator, not a blocker.

Claude Agent

coverage-auditor

Dispatched by the tdd skill (Phase 4) to audit test coverage against TDD obligations. Identifies untested source files, coverage below the maturity threshold, and logical test gaps.

Claude Agent

decision-challenger

Adversarial red-team reviewer of ADRs. Builds the strongest case against each decision, names load-bearing assumptions, assigns confidence 1–5, and surfaces evidence that would prove a decision wrong. Read-only. Dispatched optionally by decision-variance. Reads ADRs from .codearb

Claude Agent

dependency-reviewer

Dispatched when package.json, lock files, or container base images change. Verifies license, provenance, maintenance signal, and supply-chain posture against .codearbiter/security-controls.md and .codearbiter/tech-stack.md before merge.

Claude Agent

design-quality-reviewer

Reviews generated, user-facing visual or formatted output (UI, reports, slides, charts, diagrams, CLI output) against the anti-slop-design reference. Read-only; surfaces design-slop findings. Dispatched by frontend-author on UI changes. Tier 2 document producers (/pr, release) ap

Claude Agent

finding-triage

Post-processes all checkpoint reviewer reports — consolidates findings, classifies each by severity and whether it blocks the current change. Sequential. Produces a unified triage report for the checkpoint-aggregator. Reads project state from .codearbiter/.

Claude Agent

frontend-author

Use when writing or modifying frontend/UI code. Owns the TDD workflow, component conventions, state management, and UI security. MUST write failing tests before implementation code. Reads tech stack from ${CLAUDE_PROJECT_DIR}/.codearbiter/tech-stack.md.

Claude Agent

grader

INTERNAL SMARTS analyst dispatched by the decision-variance skill. Produces a SMARTS analysis and recommendation for one (artifact-position, scaffold-evidence) pair. Never decides — the user decides. Never dispatch directly.

Claude Agent

INDEX

Imported from arbiterforge/codearbiter/plugins/ca/agents/INDEX.md.

Claude Agent

infra-author

Use when writing or modifying IaC, containers, CI/CD manifests, or deployment configuration. Reads tech stack and security boundaries from ${CLAUDE_PROJECT_DIR}/.codearbiter/.

Claude Agent

map-deps

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload dependency/integration-surface mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

map-structure

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload structural mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

migration-reviewer

Dispatched when a database migration file is added or modified. Reviews migrations for safety, data-classification tagging, and immutability against .codearbiter/security-controls.md.

Claude Agent

scout

INTERNAL evidence-gatherer dispatched by the decision-variance and context-creation skills. Scans an assigned code scope and reports evidence of architectural decisions — file paths and line numbers only. Never dispatch directly.

Claude Agent

security-reviewer

Dispatch PROACTIVELY when a change touches authentication, authorization, cryptography, secrets, deployment manifests, network policies, or CI workflows. Reviews diffs against ${CLAUDE_PROJECT_DIR}/.codearbiter/security-controls.md. Read-only; produces findings.

Claude Agent

tribunal-lens-reviewer

Dispatched by the tribunal deep-audit lane, once per active lens, with a lens assignment. Generic read-only lens executor — reads the assigned lens mandate under skills/tribunal/references/lenses/ and writes one file per finding. Never dispatch directly.

Claude Agent

architecture-drift-reviewer

Read-only checkpoint reviewer. Surfaces drift between the codebase and accepted ADRs in .codearbiter/decisions/. Informational — never blocks.

Claude Agent

auth-crypto-reviewer

Reviews authentication, cryptography, key handling, and secrets against <project-root>/.codearbiter/security-controls.md. Hard blocks on banned primitives, exposed secrets, disabled TLS verification, and shell injection. Read-only checkpoint reviewer.

Claude Agent

backend-author

Use when writing or modifying backend/server-side code. Owns the TDD workflow, input validation, framework conventions, and ORM usage. MUST write failing tests before implementation code. Reads tech stack from <project-root>/.codearbiter/tech-stack.md.

Claude Agent

checkpoint-aggregator

Composes the finding-triage report and decision-challenger output into a dated checkpoint document under .codearbiter/checkpoints/YYYY-MM-DD.md. Aggregator, not a blocker.

Claude Agent

coverage-auditor

Dispatched by the tdd skill (Phase 4) to audit test coverage against TDD obligations. Identifies untested source files, coverage below the maturity threshold, and logical test gaps.

Claude Agent

decision-challenger

Adversarial red-team reviewer of ADRs. Builds the strongest case against each decision, names load-bearing assumptions, assigns confidence 1–5, and surfaces evidence that would prove a decision wrong. Read-only. Dispatched optionally by decision-variance. Reads ADRs from .codearb

Claude Agent

dependency-reviewer

Dispatched when package.json, lock files, or container base images change. Verifies license, provenance, maintenance signal, and supply-chain posture against .codearbiter/security-controls.md and .codearbiter/tech-stack.md before merge.

Claude Agent

design-quality-reviewer

Reviews generated, user-facing visual or formatted output (UI, reports, slides, charts, diagrams, CLI output) against the anti-slop-design reference. Read-only; surfaces design-slop findings. Dispatched by frontend-author on UI changes. Tier 2 document producers (/pr, release) ap

Claude Agent

finding-triage

Post-processes all checkpoint reviewer reports — consolidates findings, classifies each by severity and whether it blocks the current change. Sequential. Produces a unified triage report for the checkpoint-aggregator. Reads project state from .codearbiter/.

Claude Agent

frontend-author

Use when writing or modifying frontend/UI code. Owns the TDD workflow, component conventions, state management, and UI security. MUST write failing tests before implementation code. Reads tech stack from <project-root>/.codearbiter/tech-stack.md.

Claude Agent

grader

INTERNAL SMARTS analyst dispatched by the decision-variance skill. Produces a SMARTS analysis and recommendation for one (artifact-position, scaffold-evidence) pair. Never decides — the user decides. Never dispatch directly.

Claude Agent

INDEX

Imported from arbiterforge/codearbiter/plugins/ca-pi/agents/INDEX.md.

Claude Agent

infra-author

Use when writing or modifying IaC, containers, CI/CD manifests, or deployment configuration. Reads tech stack and security boundaries from <project-root>/.codearbiter/.

Claude Agent

map-deps

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload dependency/integration-surface mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

map-structure

Dispatched by the tribunal deep-audit lane's Phase 1, on a large/sprawling repo, to offload structural mapping out of the orchestrator's retained context. Read-only extractor, not a judge — reports facts, files no findings.

Claude Agent

migration-reviewer

Dispatched when a database migration file is added or modified. Reviews migrations for safety, data-classification tagging, and immutability against .codearbiter/security-controls.md.

Claude Agent

scout

INTERNAL evidence-gatherer dispatched by the decision-variance and context-creation skills. Scans an assigned code scope and reports evidence of architectural decisions — file paths and line numbers only. Never dispatch directly.

Claude Agent

security-reviewer

Dispatch PROACTIVELY when a change touches authentication, authorization, cryptography, secrets, deployment manifests, network policies, or CI workflows. Reviews diffs against <project-root>/.codearbiter/security-controls.md. Read-only; produces findings.

Claude Agent

tribunal-lens-reviewer

Dispatched by the tribunal deep-audit lane, once per active lens, with a lens assignment. Generic read-only lens executor — reads the assigned lens mandate under skills/tribunal/references/lenses/ and writes one file per finding. Never dispatch directly.

Claude Agent

architecture-drift-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/architecture-drift-reviewer.md.

Claude Agent

auth-crypto-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/auth-crypto-reviewer.md.

Claude Agent

backend-author

Imported from arbiterforge/codearbiter/site/src/curated/agents/backend-author.md.

Claude Agent

checkpoint-aggregator

Imported from arbiterforge/codearbiter/site/src/curated/agents/checkpoint-aggregator.md.

Claude Agent

coverage-auditor

Imported from arbiterforge/codearbiter/site/src/curated/agents/coverage-auditor.md.

Claude Agent

decision-challenger

Imported from arbiterforge/codearbiter/site/src/curated/agents/decision-challenger.md.

Claude Agent

dependency-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/dependency-reviewer.md.

Claude Agent

design-quality-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/design-quality-reviewer.md.

Claude Agent

finding-triage

Imported from arbiterforge/codearbiter/site/src/curated/agents/finding-triage.md.

Claude Agent

frontend-author

Imported from arbiterforge/codearbiter/site/src/curated/agents/frontend-author.md.

Claude Agent

grader

Imported from arbiterforge/codearbiter/site/src/curated/agents/grader.md.

Claude Agent

infra-author

Imported from arbiterforge/codearbiter/site/src/curated/agents/infra-author.md.

Claude Agent

map-deps

Imported from arbiterforge/codearbiter/site/src/curated/agents/map-deps.md.

Claude Agent

map-structure

Imported from arbiterforge/codearbiter/site/src/curated/agents/map-structure.md.

Claude Agent

migration-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/migration-reviewer.md.

Claude Agent

scout

Imported from arbiterforge/codearbiter/site/src/curated/agents/scout.md.

Claude Agent

security-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/security-reviewer.md.

Claude Agent

tribunal-lens-reviewer

Imported from arbiterforge/codearbiter/site/src/curated/agents/tribunal-lens-reviewer.md.

Commands (123)

  • /add-dep — Gates a new or changed third-party dependency through review before anything installs. Naming the source
  • /add-dep — Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs. source
  • /add-dep — Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs. source
  • /adr — Author a numbered, dated, user-attributed Architecture Decision Record under .codearbiter/decisions/. source
  • /adr — Records an architectural decision as a numbered, dated ADR under `.codearbiter/decisions/`. This is source
  • /adr — Author a numbered, dated, user-attributed Architecture Decision Record under .codearbiter/decisions/. source
  • /adr-status — Report the health of Architecture Decision Records — aged, unchallenged, supersession candidates, unresolved CONFIRM-NN. Read-only. source
  • /adr-status — Report the health of Architecture Decision Records — aged, unchallenged, supersession candidates, unresolved CONFIRM-NN. Read-only. source
  • /adr-status — A read-only health scan of every recorded ADR under `.codearbiter/decisions/`. For each one it source
  • /another — Another command, description only, no argument hint. source
  • /audit — Assembles everything codeArbiter logs — `overrides.log`, `triage.log`, `decisions/`, source
  • /audit — Assemble the governance record for a range — commits, overrides, ADRs, sprint auto-decisions, open questions, checkpoint findings — into one dated audit packet. Read-only. source
  • /audit — Assemble the governance record for a range — commits, overrides, ADRs, sprint auto-decisions, open questions, checkpoint findings — into one dated audit packet. Read-only. source
  • /btw — Lightweight Q&A about the project — answer from context and return, no routing, no state change. source
  • /btw — Lightweight Q&A about the project — answer from context and return, no routing, no state change. source
  • /btw — The one exception to codeArbiter's slash-command pipeline: a lightweight question-and-answer source
  • /checkpoint — Periodic multi-reviewer sweep of the whole codebase — surfaces a triaged checkpoint report. source
  • /checkpoint — A periodic sweep of the entire codebase with the same reviewer fleet `/ca:review` uses per-diff, source
  • /checkpoint — Periodic multi-reviewer sweep of the whole codebase — surfaces a triaged checkpoint report. source
  • /chore — Sanctioned lane for non-behavioral work — docs-only edits, dependency bumps, reverts. Type-scaled gates; no TDD demanded of prose. source
  • /chore — This is the lane for changes with no behavior to test-drive — prose edits, a version bump on an source
  • /chore — Sanctioned lane for non-behavioral work — docs-only edits, dependency bumps, reverts. Type-scaled gates; no TDD demanded of prose. source
  • /cleanup — Finish an already-merged branch — classify the leftover artifacts, return to a fast-forwarded default checkout, and delete the merged local branch. Every discard confirmed per item; containment proven, never assumed. source
  • /cleanup — Finish an already-merged branch — classify the leftover artifacts, return to a fast-forwarded default checkout, and delete the merged local branch. Every discard confirmed per item; containment proven, never assumed. source
  • /cleanup — Use this after a pull request has merged but your local checkout is still on the topic branch. source
  • /commands — Show the codeArbiter command catalog — the public command list and what each routes to. source
  • /commands — Show the codeArbiter command catalog — the public command list and what each routes to. source
  • /commands — Prints the public command catalog straight from `COMMANDS.md` — the plugin's own single source source
  • /commit — Run the full commit gate — the only sanctioned path to a git commit. source
  • /commit — This is the single entry point for turning staged work into a commit — nothing in codeArbiter source
  • /commit — Run the full commit gate — the only sanctioned path to a git commit. source
  • /conflict — The protocol for a rule conflict — not a skill route, an orchestrator-level halt. When two sources source
  • /conflict — Stop everything and surface a rule conflict — persona vs. docs vs. code. Present both sides and the conflict-hierarchy level; the user resolves. No silent reconciliation. source
  • /conflict — Stop everything and surface a rule conflict — persona vs. docs vs. code. Present both sides and the conflict-hierarchy level; the user resolves. No silent reconciliation. source
  • /context-check — An optional, on-demand drift audit for the bypass case: a merge, a direct push, or a manual edit source
  • /context-check — Optional manual drift audit — report stale provenance-tracked docs, then per stale doc offer re-scout, re-baseline, or defer. Not the daily loop; commit-gate auto-heal owns routine maintenance. source
  • /context-check — Optional manual drift audit — report stale provenance-tracked docs, then per stale doc offer re-scout, re-baseline, or defer. Not the daily loop; commit-gate auto-heal owns routine maintenance. source
  • /create-context — Brownfield back-fill — scout an existing codebase and populate .codearbiter/, then lock it initialized. source
  • /create-context — This is the populator for a project that already has code to read. Instead of interviewing you about source
  • /create-context — Brownfield back-fill — scout an existing codebase and populate .codearbiter/, then lock it initialized. source
  • /debug — Investigate-then-decide root-cause analysis for a defect whose cause is unknown. No code changes — exits to /ca:fix, /ca:adr, or a no-action close. source
  • /debug — This is where an unexplained defect goes before anyone touches code. The investigation is source
  • /debug — A fixture command named debug, colliding with the debug skill/agent. source
  • /debug — Investigate-then-decide root-cause analysis for a defect whose cause is unknown. No code changes — exits to {{CMD:fix}}, {{CMD:adr}}, or a no-action close. source
  • /decompose — This is the populator for a project that has no code yet to read. Rather than guessing at source
  • /decompose — Greenfield decomposition interview — a layered interview that populates .codearbiter/ and locks it initialized. source
  • /decompose — Greenfield decomposition interview — a layered interview that populates .codearbiter/ and locks it initialized. source
  • /doctor — Verify the active host install, package, command ownership, enforcement, and harmless live-fire probe. Read-only. source
  • /doctor — Proves the install is actually enforcing, rather than just present. codeArbiter's worst failure source
  • /doctor — Verify the active host install, package, command ownership, enforcement{{IF:pi}}, wrapper self-test, and active-dispatch coverage gap{{ELSE}}, and harmless live-fire probe{{END}}. Read-only. source
  • /feature — This is the standard entry point for new work with a human in the loop at every step. A short source
  • /feature — Start a feature: brainstorm a spec, get it approved, then drive it test-first through the pipeline. The one entry to implementation. source
  • /feature — Start a feature: brainstorm a spec, get it approved, then drive it test-first through the pipeline. The one entry to implementation. source
  • /fix — This is the entry point for a defect that already has a known cause, or one you can describe source
  • /fix — Fix a confirmed bug: a failing regression test first, then a minimal fix, then the rest of the tdd gates. source
  • /fix — Fix a confirmed bug: a failing regression test first, then a minimal fix, then the rest of the tdd gates. source
  • /init — This is how a repository opts into codeArbiter for the first time. It writes the root-level state source
  • /init — Opt this repo into codeArbiter — scaffold the root-level .codearbiter/ state store. source
  • /init — Opt this repo into codeArbiter — scaffold the root-level .codearbiter/ state store. source
  • /metrics — Read-only 3-metric governance glance — override rate, small-lane rate, sprint low-confidence ratio — each with a trend arrow vs. the prior 20-commit window. source
  • /metrics — Read-only 3-metric governance glance — override rate, small-lane rate, sprint low-confidence ratio — each with a trend arrow vs. the prior 20-commit window. source
  • /metrics — A bare-numbers governance glance — three metrics, each with a trend arrow against the prior source
  • /new-skill — The only permitted entry to creating a new codeArbiter skill. It hands off to the `skill-author` source
  • /new-skill — Author a new codeArbiter skill: prove the gap is real, get the spec approved, then write it. source
  • /new-skill — Author a new codeArbiter skill: prove the gap is real, get the spec approved, then write it. source
  • /override — Sanctioned, logged bypass of a gate or hard rule — one audit line, then proceed. source
  • /override — The sanctioned, logged escape hatch. A routine gate — a lint rule, a style check, a non-security source
  • /override — Sanctioned, logged bypass of a gate or hard rule — one audit line, then proceed. source
  • /pr — Open a pull request the only sanctioned way — clear every BLOCK-level review finding, then stage the PR. Never a direct write to the default branch. source
  • /pr — This is the only path to opening a pull request — there's no direct push or force-push to the source
  • /pr — Open a pull request the only sanctioned way — clear every BLOCK-level review finding, then stage the PR. Never a direct write to the default branch. source
  • /preview — Zero-onboarding, read-only dry-run of the reviewer fleet against the current uncommitted diff. Predicts reviewers, runs the state-free secret scan, writes nothing. source
  • /preview — Zero-onboarding, read-only dry-run of the reviewer fleet against the current uncommitted diff. Predicts reviewers, runs the state-free secret scan, writes nothing. source
  • /preview — A zero-onboarding, read-only dry-run of the reviewer fleet against whatever is currently source
  • /prune — A fixture command named prune, used to regression-test forge badges. source
  • /prune — Trim transcript clutter to extend session lifetime — analyze, prune a copy, or toggle the after-each-turn service. Dry-run by default; gains land at resume/compaction, not the current turn. source
  • /prune — Trim transcript clutter to extend session lifetime — analyze, prune a copy, or toggle the after-each-turn service. Dry-run by default; gains land at resume/compaction, not the current turn. source
  • /prune — This is a Feature Forge preview command — the after-each-turn service ships **off** by default and source
  • /reconcile — SMARTS arbitration — reconcile architectural artifacts against the scaffold and prior decisions; every variance resolved by an explicit, user-attributed choice. source
  • /reconcile — Arbitrates variances between the project's three architectural artifacts source
  • /reconcile — SMARTS arbitration — reconcile architectural artifacts against the scaffold and prior decisions; every variance resolved by an explicit, user-attributed choice. source
  • /refactor — This is the lane for moving or reshaping code without changing what it does — a rename, an extract, source
  • /refactor — Restructure code with behavioral parity proven through unmodified pre-existing tests, then refactor. No behavior change. source
  • /refactor — Restructure code with behavioral parity proven through unmodified pre-existing tests, then refactor. No behavior change. source
  • /release — Cut a release the only sanctioned way — derive the target's declared version policy from the commit log, roll its changelog, compose an annotated tag, and optionally publish its exact declared assets. Takes the declared target's name as its only argument, or --dry-run to preview one with no write. The only path to a version tag. source
  • /release — Cut a release the only sanctioned way — derive the target's declared version policy from the commit log, roll its changelog, compose an annotated tag, and optionally publish its exact declared assets. Takes the declared target's name as its only argument, or --dry-run to preview one with no write. The only path to a version tag. source
  • /release — This is the only sanctioned way a version tag gets created, and it holds no knowledge of any source
  • /review — Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR. source
  • /review — Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR. source
  • /review — A read-only review of the current change, dispatched by path against a reviewer-to-path matrix: source
  • /sample — A sample command for fixtures — single sanctioned path to a thing. source
  • /sandbox — Pull an untrusted repo into an ephemeral, host-FS-isolated Docker container — clone into a named volume, build a dep-cached image, run under structural isolation. Network defaults to offline. Requires Docker and nixpacks. source
  • /sandbox-cp — Copy a file OUT of a running sandbox box to the host — host-initiated egress only (docker cp). The reverse, a host→container bind, is impossible by construction. source
  • /sandbox-destroy — Tear down a sandbox box — remove its container and named volume. --keep-volume leaves the volume; with no id, prune reclaims any leaked ca.sandbox=1-labeled object. Cached images are retained. source
  • /sandbox-exec — Run a single command inside a running sandbox box and capture a JSON result — exitCode, separate stdout/stderr, and a truncated flag past the byte cap. The scriptable exec seam. source
  • /sandbox-shell — Open an interactive shell inside a running sandbox box at /work/repo. Read-only root, non-root user, no host-FS access — explore the untrusted code interactively, then exit. source
  • /spike — Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or /ca:feature. source
  • /spike — This is the outlet for "I need to write code to find out" rather than "I know what to build." The source
  • /spike — Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or {{CMD:feature}}. source
  • /sprint — Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops. source
  • /sprint — Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops. source
  • /sprint — This is codeArbiter's autonomy mode. It collapses the usual spec-plan-execute cycle into one source
  • /standup — Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes. source
  • /standup — Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes. source
  • /standup — The daily hygiene checklist, made routine and gated: fetch and offer a fast-forward pull, list source
  • /status — Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only. source
  • /status — A read-only snapshot of `.codearbiter/` state: the project's `stage:` maturity value, every source
  • /status — Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only. source
  • /statusline — Wire codeArbiter's statusline into ~/.claude/settings.json, or remove it. source
  • /statusline — Wires codeArbiter's renderer into `~/.claude/settings.json`, or removes it. A plugin can't own a source
  • /statusline — Wire codeArbiter's statusline into ~/.claude/settings.json, or remove it. source
  • /task — The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md. source
  • /task — The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md. source
  • /task — The only blessed way to mutate `.codearbiter/open-tasks.md`. It runs a thin writer over pure source
  • /threat-model — Opt-in lightweight STRIDE pass for a sensitive feature before implementation. Not a routine gate — invoke it when a change warrants security thought. source
  • /threat-model — Opt-in lightweight STRIDE pass for a sensitive feature before implementation. Not a routine gate — invoke it when a change warrants security thought. source
  • /threat-model — An opt-in, lightweight STRIDE pass over a design before it's built — invoked deliberately, never source
  • /tribunal — Deep, rarely-convened whole-codebase audit — eleven specialist lenses, a resumable on-disk audit log, findings filed as GitHub issues on approval. Expensive; estimates cost and STOPs before running. Never a required gate. source
  • /tribunal — codeArbiter's deepest, most expensive review, convened rarely and never as a required gate. Eleven source
  • /tribunal — Deep, rarely-convened whole-codebase audit — eleven specialist lenses, a resumable on-disk audit log, findings filed as GitHub issues on approval. Expensive; estimates cost and STOPs before running. Never a required gate. source
  • /watch — Watches a pull request's CI checks to completion without polling by hand. The wait is a real source
  • /watch — Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges. source
  • /watch — Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges. source