Autonomous Spec-Driven Development powered by GitHub SpecKit. Includes archive-aware autopilot startup, SDD coaching, multi-spec project management, and a fully autonomous workflow executor with multi-agent clarification consensus.
Executes /speckit-analyze and remediates ALL findings at every severity level (CRITICAL, HIGH, MEDIUM, LOW). After running the analysis, this agent researches each finding using web search, library docs, codebase exploration, and local file analysis to determine evidence-grounded fixes, then applies them to the relevant artifacts. Use for the analyze phase in the autopilot workflow.
Executes a single /speckit-checklist domain and remediates any [Gap] markers found. After running the checklist, this agent researches each gap using web search, library docs, codebase exploration, and local file analysis to determine evidence-grounded fixes, then applies them to spec.md or plan.md. Use for every checklist domain in the autopilot workflow.
Prepares a single Clarify question set for the autopilot workflow. This read-only agent inspects the workflow prompt, feature spec, and repo evidence, then returns prioritized questions with recommended answers and evidence for the parent orchestrator to answer/apply. It never edits artifacts and never waits on a user.
Analyzes the existing codebase to resolve questions from the perspective of established code patterns and conventions. Used by speckit-autopilot during consensus resolution for Clarify (answering questions), Checklist (remediating gaps), and Analyze (fixing findings). Spawned with a specific question, gap description, or finding — returns a structured answer with file-level evidence from the codebase.
Synthesizes outputs from the three consensus analysts (codebase-analyst, spec-context-analyst, domain-researcher) into a single actionable answer with confidence assessment. Applies the 2-of-3 agreement rule, flags all-disagree cases for human review, and produces exact artifact edits for the orchestrator to apply. Used after every consensus round in the autopilot workflow.
MANDATORY for SpecKit / Spec-Driven Development (SDD) pre-spec scoping. Use this skill — NOT brainstorming — before /speckit-specify, /speckit-plan, /speckit-tasks, /speckit-pro:speckit-scaffold-spec, or whenever the user invokes /speckit-pro:grill-me. Triggers on grill-me-unique signatures: 'grill me' on a brief/spec/transcript, 'walk every branch of the design tree', 'play the role of a relentless interviewer', 'produce a Design Concept doc', 'pre-spec scoping', 'help me scope this raw idea before /speckit-specify', 'slice-sizing', 'is this spec too big to split', 'recommend a vertical-slice split'. Walks every branch of the design tree, asks one question at a time with the assistant's recommended answer first, produces a Design Concept Markdown doc that downstream /speckit-specify, /speckit-plan, /speckit-tasks consume. Accepts .md, .txt files or a free-text topic. Use brainstorming skill ONLY for free-form creative work with no SpecKit/SDD anchor.
Archive a merged SpecKit spec and clean active workflow residue after the implementation PR has merged. Use after confirming merge provenance, when the user asks for post-merge SpecKit archive hygiene, cleanup hygiene, or removal of completed specs from active specs.
Autonomous SpecKit workflow executor. Reads a populated workflow file and executes all 7 SDD phases (specify → clarify → plan → checklist → tasks → analyze → implement) with programmatic gate validation, multi-agent consensus resolution, and auto-commits. Use when the user says "run autopilot", "execute workflow", "autonomous speckit", or has a workflow file ready for execution.
Coaches developers through Spec-Driven Development using the official SpecKit CLI and the speckit-pro plugin. Provides SDD methodology guidance, per-command coaching, phase gate validation, multi-spec technical roadmap creation, workflow tracking, existing-project fixup, and live SpecKit extension and preset discovery. Use when user asks about SDD methodology, SpecKit commands, spec writing, feature decomposition, gate failures, technical roadmaps, the roadmap map / home note / Map of Content / navigation and its curated-vs-generated two-zone structure, autopilot workflows, project repair, template durability, consensus protocol, or browsing, installing, configuring, and removing SpecKit community extensions, presets, hooks, and the catalog.
Installs the official SpecKit CLI and initializes one or both coding-agent integrations (Claude Code, Codex CLI). Detects existing installs and hands off to /speckit-pro:speckit-upgrade rather than overwriting. Optionally installs the curated set of community extensions and presets. Use when the user says "install speckit", "set up speckit", "initialize speckit", "add speckit to this repo", "install spec-kit", "bootstrap speckit", "first-time speckit setup", "install the specify cli", "set up specify", or wants to install for claude only, codex only, or both side-by-side. Not for upgrading an existing install (use /speckit-pro:speckit-upgrade) or running workflows (use /speckit-pro:speckit-autopilot).
Uses power tools
Uses Bash, Write, or Edit tools
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 claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Racecraft Plugins Public is the open-source plugin marketplace from Racecraft Lab. It currently publishes SpecKit Pro for both Claude Code and Codex.
The short version: install this marketplace if you want Racecraft-maintained agent workflows you can inspect, version, and update through your coding-agent plugin system.
Trust check: Plugins can change how an agent reads files, writes files, launches tools, or delegates work. Review the plugin README and source before installing or updating.
| You are... | Install if... | Skip for now if... |
|---|---|---|
| Evaluating Racecraft plugins | You want to inspect and install public Racecraft plugins from one place. | You only need a one-off prompt or local personal workflow. |
| Using Spec-Driven Development | You want a guided Spec Kit workflow with scoping, planning, gates, and implementation support. | Your change is tiny enough for a direct manual edit. |
| Maintaining plugins | You want examples of dual Claude Code and Codex plugin packaging. | You do not need marketplace distribution or generated payloads. |
| Plugin | What it does | Best first step |
|---|---|---|
| SpecKit Pro | Adds Racecraft workflows around GitHub Spec Kit: coaching, scoping interviews, spec scaffolding, autopilot execution, status, and PR review resolution. | Read the SpecKit Pro decision guide. |
This repository exposes the same marketplace through two runtime surfaces:
| Runtime | Marketplace metadata | Install payload |
|---|---|---|
| Claude Code | .claude-plugin/marketplace.json | dist/claude/speckit-pro |
| Codex | .agents/plugins/marketplace.json | dist/codex/speckit-pro |
Add the marketplace, then install the plugin:
/plugin marketplace add racecraft-lab/racecraft-plugins-public
/plugin install speckit-pro@racecraft-plugins-public
/reload-plugins
Verify the installed plugin from Claude Code's /plugin UI, then use
namespaced plugin skills such as /speckit-pro:speckit-status and
/speckit-pro:speckit-coach.
Claude Code commands are the separate DOC-003-owned path; use the Claude Code install guide for that runtime. For Codex, open this repository in Codex, then open the plugin directory:
codex
/plugins
Codex reads the repo-scoped marketplace from
.agents/plugins/marketplace.json. That
marketplace points SpecKit Pro at the generated Codex payload,
dist/codex/speckit-pro/. Do not install Codex
from the mixed authoring tree at speckit-pro/.
After installing SpecKit Pro, run the plugin's Codex install skill so its bundled custom-agent TOML files are copied into Codex's agent registry:
@SpecKit Pro -> install
or:
$install
The default destination is ~/.codex/agents/; choose .codex/agents/ only when
you intentionally want project-scoped custom agents. Verify these ten
installer-copied TOML files only:
autopilot-fast-helper.tomlphase-executor.tomlclarify-executor.tomlchecklist-executor.tomlanalyze-executor.tomlimplement-executor.tomlcodebase-analyst.tomlspec-context-analyst.tomldomain-researcher.tomluat-runbook-author.tomlThen restart Codex. Also restart Codex after plugin enablement changes, custom-agent refreshes, or relevant Codex config edits.
Codex loads installed plugins from the installed plugin cache at
~/.codex/plugins/cache/$MARKETPLACE_NAME/$PLUGIN_NAME/$VERSION/. Treat that as
runtime state, not the source of truth. If SpecKit Pro looks stale after an
update, check the marketplace source or copied personal payload, the generated
payload, the installed plugin cache, the selected custom-agent destination, and
whether Codex was restarted. Rerun @SpecKit Pro -> install or $install after
a plugin update that changes bundled custom-agent TOML files, then restart
Codex again. Do not edit the installed plugin cache.
Code intelligence powered by a knowledge graph. Provides execution flow tracing, blast radius analysis, and augmented search across your codebase.
npx claudepluginhub racecraft-lab/racecraft-plugins-public --plugin speckit-proHarness-native ECC operator layer - 67 agents, 278 skills, 94 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
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.
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.
Matt Pocock's agent skills for real engineering — grilling, spec/ticket flows, TDD, code review, domain modelling and more. Plug-and-play, not vibe coding.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques