By JackReis
Engineering lifecycle skills forked from Matt Pocock's framework: SDLC scaffolding (setup-matt-pocock-skills), test-driven development (tdd), plan-to-issues decomposition (to-issues), context-to-PRD (to-prd), issue triage state machine (triage), disciplined diagnosis (diagnose), architecture deepening (improve-codebase-architecture), and zoom-out perspective (zoom-out). Run setup-matt-pocock-skills first to configure your repo before using the others.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.
Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.
Sets up an `## Agent skills` block in AGENTS.md/CLAUDE.md and `docs/agents/` so the engineering skills know this repo's issue tracker (GitHub or local markdown), triage label vocabulary, and domain doc layout. Run before first use of `to-issues`, `to-prd`, `triage`, `diagnose`, `tdd`, `improve-codebase-architecture`, or `zoom-out` — or if those skills appear to be missing context about the issue tracker, triage labels, or domain docs.
Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.
Break a plan, spec, or PRD into independently-grabbable issues on the project issue tracker using tracer-bullet vertical slices. Use when user wants to convert a plan into issues, create implementation tickets, or break down work into issues.
A2A-native agent consensus framework — 5 packs, 26 skills, zero inherited bloat.
Arbiter is an A2A-native agent consensus framework delivered as a curated collection of Claude Code skill packs. Agents design, reconcile, ratify, and audit — with filesystem durability as the floor and JSON-RPC/SSE as the live wire.
Originally derived from an MIT-licensed upstream, inherited plugins were removed on 2026-05-16. Arbiter now ships only original contributions.
| Pack | Category | Skills | Description |
|---|---|---|---|
autonomous-ai-agents v0.4.0 | ai-agency | 3 + 2 MCP bridges | Fleet coordination: identity, hermes-bridge, openclaw-bridge |
grill-each-other v1.3.1 | skill-enhancers | 10 | Dialectic claim discipline: grill-me, peer-grill, fleet-ratify, permutation, etc. |
athenaeum v0.2.0 | skill-enhancers | 4 | Streamlined dialectic: design, reconcile, ratify, audit |
leonardo v1.1.0 | ai-agency | 1 | Protected-string decoder with audit trail |
pocock-engineering v1.0.0 | skill-enhancers | 8 | SDLC skills forked from Matt Pocock's framework |
| Skill | What it does |
|---|---|
agent-identity | Declares who an agent is — name, role, capabilities, boundaries |
hermes-bridge | MCP bridge to Hermes messaging relay |
openclaw-bridge | MCP bridge to OpenClaw dispatch system |
| Skill | What it does |
|---|---|
grill-me | Interview the user about a plan until shared understanding |
grill-me-agents | Grill multi-agent designs — roles, handoffs, failure modes |
grill-me-with-agents | Code-aware variant that cross-references existing agent definitions |
grill-with-docs | Grill against domain model and documented decisions |
peer-grill | Two agents interrogate each other to converge on shared state |
agent-show-and-tell | Agents write status reports; one reader collates a roundup |
fleet-ratify | Ratify fleet decisions with SHA-256 attestation |
permutation | NxN fleet topology ratification with visual diagrams |
caveman | Ultra-compressed communication mode (75% token reduction) |
find-skills | Discover and install agent skills |
| Skill | What it does |
|---|---|
design | Propose a design claim for peer review |
reconcile | Resolve conflicting claims into a merged design |
ratify | Sign off on a finalized design with attestation |
audit | 13-branch audit protocol with confidence rules |
| Skill | What it does |
|---|---|
protected-string-decoder | Decode protected strings with audit trail |
| Skill | What it does |
|---|---|
zoom-out | Step back for broader architectural context |
diagnose | Reproduce → minimize → hypothesize → instrument → fix → regression-test |
triage | State-machine issue triage with role-based workflows |
tdd | Test-driven development with red-green-refactor |
to-issues | Break plans into tracer-bullet vertical-slice issues |
to-prd | Turn conversation context into a PRD on the issue tracker |
improve-codebase-architecture | Find refactoring opportunities informed by domain language |
setup-matt-pocock-skills | Bootstrap AGENTS.md with issue tracker, triage labels, domain docs |
# Install a pack
/plugin install grill-each-other@arbiter
/plugin install athenaeum@arbiter
Or install individual skills by copying skills/<name>/SKILL.md into your own project.
All skills follow the 2025 schema — name + description as the portable minimum, with allowed-tools, version, and package manifests as optional adapter fields.
See AGENTS.md for the full developer guide, schema details, and GitNexus integration.
Arbiter speaks the Agent-to-Agent Protocol natively:
agent_card.py generates capability descriptors with Athenaeum extensionsa2a_task.py serializes design/reconcile/ratify/audit workflows to JSONtasks/send, tasks/get, tasks/cancel on port 18765tasks/sendSubscribe pushes live task updates until terminal stateathenaeum poll <topic> for zero-server async status checksA2A is opt-in. Default mode is filesystem-only:
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub jackreis/arbiterCreate ADHD-friendly visual organizational tools (Mermaid diagrams) optimized for neurodivergent thinking patterns with accessibility modes
Engineering lifecycle skills forked from Matt Pocock's framework: SDLC scaffolding (setup-matt-pocock-skills), test-driven development (tdd), plan-to-issues decomposition (to-issues), context-to-PRD (to-prd), issue triage state machine (triage), disciplined diagnosis (diagnose), architecture deepening (improve-codebase-architecture), and zoom-out perspective (zoom-out). Run setup-matt-pocock-skills first to configure your repo before using the others.
Dialectic claim discipline for AI agents: scholastic vocabulary, peer-grill file-based reconciliation, runnable verifiers/falsifiers, multi-agent ratification, fleet-ratify N-agent attestation, permutation NxN fleet topology ratification. Synthesis: an unexamined claim does not exist.
Encode or decode mirror-scripted protected strings (left-hand write + right-hand read) with Discord audit signals (tattle) on every operation.
Minimal enclosure for gathering and sharing with the fleet — VS Code Agents plugin
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Lazy senior dev mode. Forces the simplest, shortest solution that actually works: YAGNI, stdlib first, no unrequested abstractions.
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification