speckit-pro
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.
By Racecraft LabLicense: MIT5 GitHub starsUpdated 3 days ago
Directory evidence
- Runtimes
- Claude Code
- Parsed components
- 14 skill or MCP entries
- Source updated
- Oct 4, 2026
- Manifest status
- Canonical path parsed
The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →
Install speckit-pro for Claude Code
claude plugin marketplace add IchenDEV/agent-plugin-mkt
claude plugin marketplace update agent-plugin-marketplace
claude plugin install speckit-pro@agent-plugin-marketplacePaste and run these commands in a terminal with Claude Code. They add and refresh the PluginsMP catalog, then install this plugin.
The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.
Get the source manually
git clone https://github.com/racecraft-lab/racecraft-plugins-publicClone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is dist/claude/speckit-pro/.
Plugin files
├── .claude-plugin/plugin.json├── skills/grill-me/SKILL.md├── skills/speckit-archive-cleanup/SKILL.md├── skills/speckit-autopilot/SKILL.md├── skills/speckit-coach/SKILL.md├── skills/speckit-install/SKILL.md├── skills/speckit-prd/SKILL.md├── skills/speckit-resolve-pr/SKILL.md├── skills/speckit-scaffold-spec/SKILL.md├── skills/speckit-status/SKILL.md├── skills/speckit-upgrade/SKILL.md├── skills/ubiquitous-language/SKILL.md└── .mcp.json
Included Skills11
Run an interactive, one-question-at-a-time design interview before SpecKit specification work and produce a Design Concept record. Use when an active user requests Grill Me—for example “grill me,” “interview me about,” “walk the design tree,” or “produce a Design Concept”—or invokes /speckit-pro:grill-me, or when interactive /speckit-pro:speckit-scaffold-spec delegates its required interview. SPEC setup, worktree creation, and workflow population belong to /speckit-pro:speckit-scaffold-spec; a setup request alone is not an interview delegation. SDD methodology, checklist selection, and gate guidance without a requested design interview belong to /speckit-pro:speckit-coach. Recommend one grounded answer first for every consequential choice. Not for autonomous, background, CI, autopilot, or subagent execution.
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. Only for SpecKit spec archives: do NOT use for generic git housekeeping such as pruning merged branches, deleting stale worktrees, git gc, or clearing build output, and not to scaffold a spec (use speckit-scaffold-spec), run a workflow (use speckit-autopilot), read status (use speckit-status), or fix PR review comments (use speckit-resolve-pr).
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.
Coach developers through Spec-Driven Development and SpecKit Pro. Use for SDD methodology, SpecKit command and gate guidance, technical-roadmap and workflow design, roadmap-MOC guidance, checklist selection, SpecKit project repair, selective formal-methods guidance and model/checker selection, SpecKit consensus protocol and voting rules, or SpecKit preset and extension discovery and configuration. Not for running autopilot, conducting grill-me, or unrelated coding (including MCP tool implementation). Formal-check execution and autopilot resume belong to speckit-autopilot.
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).
Collaboratively create or update a lean PRD, its 1:1 technical-roadmap SPEC catalog, and a roadmap-MOC home note. Use when an active user asks to write a PRD, create a product requirements document, turn an idea or brief into a PRD and roadmap, plan a product, or invokes /speckit-pro:speckit-prd. Read-only workflow status and next-spec recommendations belong to /speckit-pro:speckit-status. Roadmap-only decomposition or dependency planning from an existing PRD, without creating or updating the PRD, belongs to /speckit-pro:speckit-coach. Ask one grounded decision at a time, then hand off the resulting roadmap to /speckit-pro:speckit-scaffold-spec. Not for per-spec scoping, worktree preparation, or general SDD coaching.
MANDATORY for resolving GitHub PR review comments by editing the source code those comments flagged. Use this skill — NOT a read-only PR review skill — whenever the user wants to ACT on PR review feedback by changing code, committing, and pushing. Triggers on these phrases: 'resolve PR review comments', 'address review feedback', 'fix the copilot comments', 'resolve the threads on PR #N', 'fix each review comment and resolve the threads', 'handle the Copilot review on this PR', 'work through the review comments on this PR', 'address them all', 'take care of the outstanding review feedback', or whenever a PR URL is pasted with unresolved comments. The skill edits files, runs project verification, commits the fixes, pushes, posts a reply per thread, and marks each thread resolved via gh API. NOT for read-only PR review, summarizing what a PR changes, or assessing PR merge risk — those are read-only review skills, this is a write skill that mutates the working tree.
Use this skill when the user wants to set up, scaffold, bootstrap, prep, initialize, or prepare a SPEC-ID from the technical roadmap for autonomous execution. Triggers on: set up SPEC-XXX, scaffold SPEC-XXX, bootstrap SPEC-XXX for development, prep SPEC-XXX, initialize the workspace for SPEC-XXX, prepare SPEC-XXX for the autonomous run, create a spec branch and workflow for SPEC-XXX, generate the workflow file for SPEC-XXX, I need a workflow file generated for SPEC-XXX, fill the prompts from the roadmap, pre-fill the workflow template, start working on SPEC-XXX, populate the workflow file for SPEC-XXX. Opens with a blind-spot pass, creates the git worktree, spec branch, Design Concept doc, and populated workflow file, then hands off to planning. Accepts --answers-file for unattended setup; otherwise interviews the user. Not for checking roadmap status (use /speckit-pro:speckit-status), running a populated workflow (use /speckit-pro:speckit-autopilot), or SDD coaching (use /speckit-pro:speckit-coach).
Use when the user wants to read SpecKit project status, see the roadmap dashboard, check which specs are complete, in-progress, blocked, or ready to start, see phase-level progress for active specs, find which active worktrees exist and what spec each belongs to, check the status or phase of a specific SPEC-ID, identify what is blocking a spec and why, get a recommendation for the next spec to implement, or summarize overall project health and next steps. Triggers on: show roadmap, project status, active workflows, blockers, dashboard, all specs, phases complete, next phase, list worktrees, which spec should I start, what is next, SpecKit progress, project health, check SPEC-XXX, is SPEC-XXX done, is SPEC-XXX blocked. Read-only: never creates files, branches, or worktrees. Do NOT use to set up or scaffold a spec (use speckit-scaffold-spec), execute a workflow autonomously (use speckit-autopilot), fix PR review comments (use speckit-resolve-pr), or ask about SDD or SpecKit consensus/voting (use speckit-coach).
Upgrades or migrates an existing SpecKit installation safely with backup-and-restore for locally-modified files. Preserves the project constitution and template overrides. Supports upgrading one or both integrations (Claude Code, Codex CLI) and offering missing curated community extensions and presets. Use when the user asks to execute an upgrade or migration, including "upgrade speckit", "update speckit", "refresh speckit", "new speckit version", "latest speckit", "upgrade specify cli", "safe speckit upgrade", "speckit migration to skills", or "preserve my constitution during upgrade". Not for pre-upgrade project, template, or preset repair (use /speckit-pro:speckit-coach). Hands off to /speckit-pro:speckit-install if .specify/ is missing.
Build or refresh the repository's ubiquitous-language terms document, a committed Markdown table of domain terms (term, meaning here, identifiers) that humans and agents read before designing. Use when the user asks to define domain terms, create or update a glossary or ubiquitous language, pin what a term means in this codebase, reconcile names between the spec and the code, or invokes /speckit-pro:ubiquitous-language. Also runs the advisory identifier lint on a diff. Not for scoping interviews, PRD authoring, or general SDD coaching.
MCP servers3
- command
- python3
- args
- -m speckit_pro_runner.author_broker
- env.PYTHONPATH
- ${CLAUDE_PLUGIN_ROOT}
- command
- python3
- args
- -m speckit_pro_runner.research_broker
- env.PYTHONPATH
- ${CLAUDE_PLUGIN_ROOT}
- command
- python3
- args
- -m speckit_pro_runner.sweep_broker
- env.PYTHONPATH
- ${CLAUDE_PLUGIN_ROOT}
MCP configuration uses runtime-provided plugin path placeholders such as ${PLUGIN_ROOT} or ${CLAUDE_PLUGIN_ROOT}. Review the manifest for the runtime-specific expansion rules.
Plugin manifests1
{
"name": "speckit-pro",
"description": "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.",
"author": {
"name": "Racecraft Lab"
},
"homepage": "https://github.com/racecraft-lab/racecraft-plugins-public/tree/main/speckit-pro",
"license": "MIT",
"keywords": [
"speckit",
"sdd",
"spec-driven-development",
"specification",
"planning",
"autopilot",
"autonomous",
"workflow"
],
"repository": "https://github.com/racecraft-lab/racecraft-plugins-public",
"dependencies": [
"typesafe-jev"
]
}For maintainers
If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.
[speckit-pro on Agent Plugins Marketplace](https://pluginsmp.com/plugins/speckit-pro)