bigpowers 2.20.0 → 2.22.0

package/.pi/package.json

@@ -1,6 +1,6 @@
 {
   "name": "bigpowers",
-  "version": "2.20.0",
+  "version": "2.22.0",
   "description": "68 skills — 61 agent skills for spec-driven, test-first software development by solo developers",
   "keywords": [
     "pi-package"

package/.pi/prompts/build-epic.md

@@ -24,7 +24,7 @@ Orchestrates the **build** flow for a single epic: survey → plan tasks → kic
 | 5 | `verify-work` — UAT + mechanical gates |
 | 6 | `audit-code` — **non-optional gate** (pass/fail; fail → loop back to step 4) |
 | 7 | `commit-message` — Conventional Commits draft |
-| 8 | `release-branch` — PR or solo land |
+| 8 | `release-branch` — PR or solo land (supports `--squash-state`) |
 
 ## Process
 

package/.pi/prompts/change-request.md

@@ -37,6 +37,18 @@ See [REFERENCE.md](REFERENCE.md) for the full WSJF scoring rubric.
 
 → verify: `grep -c 'wsjf' specs/release-plan.yaml specs/epics/*/epic.yaml`
 
+## Conversational Mode
+
+If the user's request is in natural language and does not match the structured format of Mode A or Mode B, enter Conversational Mode to extract the change parameters through interactive dialogue.
+
+### 5-Step Flow
+
+1. **Capture**: Parse the natural-language request for what, why, and where. Ask at most 3 clarifying questions before drafting.
+2. **Locate**: Identify which epic or capability in `specs/epics/` the request affects or replaces.
+3. **Draft**: Present a structured draft of the proposed story and tasks for user confirmation.
+4. **Score**: Estimate the WSJF score and explain the calculation so the user understands the priority.
+5. **Place**: Confirm the final epic placement with the user before writing files.
+
 ## After either mode
 
 Run `bash scripts/sync-status-from-epics.sh`. Suggest `plan-work` or `build-epic` for the top-ranked unstarted story.

package/.pi/prompts/release-branch.md

@@ -12,6 +12,7 @@ Finalize a completed feature branch: verify coverage gates, integrate onto `main
 ## Additional modes
 
 - `--hotfix`: Emergency fix. Cherry-pick to main plus immediate tag. Skip PR in solo profile.
+- `--squash-state`: Squashes all intermediate `chore(state):` commits on the feature branch into a single clean commit before merging. Use this to reduce noise in the main git repository history.
 
 ## Integrate mode
 

package/.pi/skills/build-epic/SKILL.md

@@ -26,7 +26,7 @@ Orchestrates the **build** flow for a single epic: survey → plan tasks → kic
 | 5 | `verify-work` — UAT + mechanical gates |
 | 6 | `audit-code` — **non-optional gate** (pass/fail; fail → loop back to step 4) |
 | 7 | `commit-message` — Conventional Commits draft |
-| 8 | `release-branch` — PR or solo land |
+| 8 | `release-branch` — PR or solo land (supports `--squash-state`) |
 
 ## Process
 

package/.pi/skills/change-request/SKILL.md

@@ -39,6 +39,18 @@ See [REFERENCE.md](REFERENCE.md) for the full WSJF scoring rubric.
 
 → verify: `grep -c 'wsjf' specs/release-plan.yaml specs/epics/*/epic.yaml`
 
+## Conversational Mode
+
+If the user's request is in natural language and does not match the structured format of Mode A or Mode B, enter Conversational Mode to extract the change parameters through interactive dialogue.
+
+### 5-Step Flow
+
+1. **Capture**: Parse the natural-language request for what, why, and where. Ask at most 3 clarifying questions before drafting.
+2. **Locate**: Identify which epic or capability in `specs/epics/` the request affects or replaces.
+3. **Draft**: Present a structured draft of the proposed story and tasks for user confirmation.
+4. **Score**: Estimate the WSJF score and explain the calculation so the user understands the priority.
+5. **Place**: Confirm the final epic placement with the user before writing files.
+
 ## After either mode
 
 Run `bash scripts/sync-status-from-epics.sh`. Suggest `plan-work` or `build-epic` for the top-ranked unstarted story.

package/.pi/skills/release-branch/SKILL.md

@@ -14,6 +14,7 @@ Finalize a completed feature branch: verify coverage gates, integrate onto `main
 ## Additional modes
 
 - `--hotfix`: Emergency fix. Cherry-pick to main plus immediate tag. Skip PR in solo profile.
+- `--squash-state`: Squashes all intermediate `chore(state):` commits on the feature branch into a single clean commit before merging. Use this to reduce noise in the main git repository history.
 
 ## Integrate mode
 

package/CHANGELOG.md

@@ -1,3 +1,17 @@
+# [2.22.0](https://github.com/danielvm-git/bigpowers/compare/v2.21.0...v2.22.0) (2026-06-22)
+
+
+### Features
+
+* **mcp-server:** add MCP server exposing 68 skills as MCP tools ([58c4592](https://github.com/danielvm-git/bigpowers/commit/58c45925d63c505c8eadf1443b6268f4d8791715))
+
+# [2.21.0](https://github.com/danielvm-git/bigpowers/compare/v2.20.0...v2.21.0) (2026-06-22)
+
+
+### Features
+
+* **change-request:** add conversational intake path ([285cecf](https://github.com/danielvm-git/bigpowers/commit/285cecff204273f9333bb6aac6d3c378cb1209b0))
+
 # [2.20.0](https://github.com/danielvm-git/bigpowers/compare/v2.19.0...v2.20.0) (2026-06-21)
 
 

package/CONVENTIONS.md

@@ -24,6 +24,7 @@ All changes to this repository MUST follow the [Conventional Commits 1.0.0](http
 - Use `gh run view` / `gh run watch` for CI status
 - Verify auth with `gh auth status` before operations
 - **Git Attribution:** NEVER include `Co-authored-by`, `Co-Authored-By`, or any other footer that attributes code to an AI agent (e.g., Claude, Gemini). All commits must appear as if they were authored solely by the human user.
+- **State Commit Policy:** To minimize git history noise, intermediate `chore(state):` commits (e.g., tracking build-epic step transitions) should either be squashed locally using the `--squash-state` flag on `release-branch` before merging, or kept out-of-band using local cycle-state files.
 - Never call GitHub REST API directly (curl, fetch, etc.)
 - Never create GitHub issues from automated workflows — produce local .md files in specs/ instead
 

package/README.md

@@ -122,6 +122,47 @@ pi install npm:bigpowers
 
 Skills are loaded on-demand via progressive disclosure: only descriptions are always in context; the full SKILL.md loads when the agent reads it. Prompt templates expand in pi's editor with autocomplete.
 
+## 🔧 MCP Server (Model Context Protocol)
+
+bigpowers ships an MCP server that exposes all skills as callable MCP tools. Agents can discover and invoke skills dynamically instead of relying on a static system prompt.
+
+### Start the server
+
+```bash
+node scripts/mcp-server.js
+```
+
+### Add to Claude Code
+
+```bash
+claude mcp add bigpowers node /path/to/bigpowers/scripts/mcp-server.js
+```
+
+Or add manually to `.claude/settings.json`:
+
+```json
+{
+  "mcpServers": {
+    "bigpowers": {
+      "command": "node",
+      "args": ["/path/to/bigpowers/scripts/mcp-server.js"]
+    }
+  }
+}
+```
+
+### Available MCP tools
+
+| Tool | Description |
+|------|-------------|
+| `bigpowers_list_skills` | List all 68 skills with name, description, phase. Optional `phase` filter. |
+| `bigpowers_get_skill` | Get full SKILL.md content for any skill by name. |
+| `bigpowers_search_skills` | Keyword/semantic search — returns ranked matches for a query. |
+| `bigpowers_get_state` | Get current `specs/state.yaml` (active flow, epic, step). |
+| `bigpowers_invoke_skill` | Get skill instructions with optional context for agent invocation. |
+
+---
+
 ## 🏗 The v2.0.0 Lifecycle
 
 Every project follows the **orchestrate-project 6-phase model** (full SOP: [`docs/WORKFLOW-SOP-v2.md`](docs/WORKFLOW-SOP-v2.md)):

package/SKILL-INDEX.md

@@ -3,7 +3,7 @@
 > **DO NOT EDIT** — This file is auto-generated by `scripts/generate-skill-index.sh`.
 > Edit `SKILL.md` source files or `skills-lock.json` instead. Run `bash scripts/sync-skills.sh` to regenerate.
 
-**Generated:** 2026-06-21T22:00:07Z
+**Generated:** 2026-06-22T02:11:51Z
 **Skills:** 68
 
 ---

package/build-epic/SKILL.md

@@ -25,7 +25,7 @@ Orchestrates the **build** flow for a single epic: survey → plan tasks → kic
 | 5 | `verify-work` — UAT + mechanical gates |
 | 6 | `audit-code` — **non-optional gate** (pass/fail; fail → loop back to step 4) |
 | 7 | `commit-message` — Conventional Commits draft |
-| 8 | `release-branch` — PR or solo land |
+| 8 | `release-branch` — PR or solo land (supports `--squash-state`) |
 
 ## Process
 

package/change-request/SKILL.md

@@ -38,6 +38,18 @@ See [REFERENCE.md](REFERENCE.md) for the full WSJF scoring rubric.
 
 → verify: `grep -c 'wsjf' specs/release-plan.yaml specs/epics/*/epic.yaml`
 
+## Conversational Mode
+
+If the user's request is in natural language and does not match the structured format of Mode A or Mode B, enter Conversational Mode to extract the change parameters through interactive dialogue.
+
+### 5-Step Flow
+
+1. **Capture**: Parse the natural-language request for what, why, and where. Ask at most 3 clarifying questions before drafting.
+2. **Locate**: Identify which epic or capability in `specs/epics/` the request affects or replaces.
+3. **Draft**: Present a structured draft of the proposed story and tasks for user confirmation.
+4. **Score**: Estimate the WSJF score and explain the calculation so the user understands the priority.
+5. **Place**: Confirm the final epic placement with the user before writing files.
+
 ## After either mode
 
 Run `bash scripts/sync-status-from-epics.sh`. Suggest `plan-work` or `build-epic` for the top-ranked unstarted story.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "bigpowers",
-  "version": "2.20.0",
+  "version": "2.22.0",
   "description": "61 agent skills for spec-driven, test-first software development by solo developers",
   "main": "index.js",
   "scripts": {

package/release-branch/SKILL.md

@@ -13,6 +13,7 @@ Finalize a completed feature branch: verify coverage gates, integrate onto `main
 ## Additional modes
 
 - `--hotfix`: Emergency fix. Cherry-pick to main plus immediate tag. Skip PR in solo profile.
+- `--squash-state`: Squashes all intermediate `chore(state):` commits on the feature branch into a single clean commit before merging. Use this to reduce noise in the main git repository history.
 
 ## Integrate mode
 

package/scripts/mcp-server.js

@@ -0,0 +1,246 @@
+#!/usr/bin/env node
+// bigpowers MCP server — exposes 68 skills as MCP tools via stdio transport
+// Zero dependencies — raw JSON-RPC over stdio with Content-Length framing
+
+'use strict';
+
+const fs = require('fs');
+const path = require('path');
+
+const ROOT = path.resolve(__dirname, '..');
+const SKILLS_LOCK = path.join(ROOT, 'skills-lock.json');
+const STATE_YAML = path.join(ROOT, 'specs', 'state.yaml');
+
+// Phase map — mirrors generate-skill-index.sh PHASE_MAP
+const PHASE_MAP = {
+  'survey-context': 'Discover', 'research-first': 'Discover', 'search-skills': 'Discover',
+  'using-bigpowers': 'Discover', 'map-codebase': 'Discover', 'elaborate-spec': 'Discover',
+  'model-domain': 'Design', 'define-language': 'Design', 'grill-me': 'Design',
+  'grill-with-docs': 'Design', 'deepen-architecture': 'Design', 'design-interface': 'Design',
+  'define-success': 'Design',
+  'scope-work': 'Plan', 'slice-tasks': 'Plan', 'plan-work': 'Plan', 'plan-release': 'Plan',
+  'plan-refactor': 'Plan', 'assess-impact': 'Plan', 'change-request': 'Plan',
+  'run-planning': 'Plan', 'seed-conventions': 'Plan',
+  'develop-tdd': 'Build', 'kickoff-branch': 'Build', 'execute-plan': 'Build',
+  'build-epic': 'Build', 'spike-prototype': 'Build', 'craft-skill': 'Build',
+  'quick-fix': 'Build', 'setup-environment': 'Build', 'wire-observability': 'Build',
+  'wire-ci': 'Build', 'publish-package': 'Build', 'align-grid': 'Build',
+  'orchestrate-project': 'Build', 'guard-git': 'Build', 'hook-commits': 'Build',
+  'deploy': 'Build', 'smoke-test': 'Build', 'validate-contracts': 'Build',
+  'verify-work': 'Verify', 'validate-fix': 'Verify', 'audit-code': 'Verify',
+  'enforce-first': 'Verify', 'run-evals': 'Verify', 'investigate-bug': 'Verify',
+  'diagnose-root': 'Verify', 'fix-bug': 'Verify', 'inspect-quality': 'Verify',
+  'request-review': 'Verify', 'respond-review': 'Verify', 'trace-requirement': 'Verify',
+  'release-branch': 'Release', 'commit-message': 'Release',
+  'session-state': 'Sustain', 'terse-mode': 'Sustain', 'compose-workflow': 'Sustain',
+  'delegate-task': 'Sustain', 'dispatch-agents': 'Sustain', 'edit-document': 'Sustain',
+  'evolve-skill': 'Sustain', 'migrate-spec': 'Sustain', 'organize-workspace': 'Sustain',
+  'reset-baseline': 'Sustain', 'simulate-agents': 'Sustain', 'stocktake-skills': 'Sustain',
+  'write-document': 'Sustain',
+};
+
+function loadSkillsCatalog() {
+  const lock = JSON.parse(fs.readFileSync(SKILLS_LOCK, 'utf8'));
+  return Object.entries(lock.skills).map(([name, meta]) => ({
+    name,
+    description: meta.description,
+    path: meta.path,
+    phase: PHASE_MAP[name] || 'Other',
+  }));
+}
+
+function getSkillContent(skillName) {
+  const catalog = loadSkillsCatalog();
+  const skill = catalog.find(s => s.name === skillName);
+  if (!skill) return null;
+  const skillPath = path.join(ROOT, skill.path);
+  if (!fs.existsSync(skillPath)) return null;
+  return fs.readFileSync(skillPath, 'utf8');
+}
+
+function searchSkills(query) {
+  const terms = query.toLowerCase().split(/\s+/).filter(Boolean);
+  const catalog = loadSkillsCatalog();
+  return catalog
+    .map(skill => {
+      const haystack = `${skill.name} ${skill.description} ${skill.phase}`.toLowerCase();
+      const score = terms.reduce((s, t) => s + (haystack.includes(t) ? 1 : 0), 0);
+      return { skill, score };
+    })
+    .filter(({ score }) => score > 0)
+    .sort((a, b) => b.score - a.score)
+    .map(({ skill }) => skill);
+}
+
+// Tool handlers
+const TOOLS = {
+  bigpowers_list_skills: {
+    description: 'List all bigpowers skills with name, description, and lifecycle phase. Optionally filter by phase.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        phase: {
+          type: 'string',
+          description: 'Filter by lifecycle phase: Discover, Design, Plan, Build, Verify, Release, Sustain',
+        },
+      },
+    },
+    handler({ phase } = {}) {
+      let skills = loadSkillsCatalog();
+      if (phase) skills = skills.filter(s => s.phase.toLowerCase() === phase.toLowerCase());
+      return skills.map(({ name, description, phase: p }) => ({ name, description, phase: p }));
+    },
+  },
+
+  bigpowers_get_skill: {
+    description: 'Get the full SKILL.md content for a specific bigpowers skill by name.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        name: { type: 'string', description: 'Skill name (kebab-case, e.g. "survey-context")' },
+      },
+      required: ['name'],
+    },
+    handler({ name }) {
+      const content = getSkillContent(name);
+      if (!content) return { error: `Skill "${name}" not found` };
+      return { name, content };
+    },
+  },
+
+  bigpowers_search_skills: {
+    description: 'Search bigpowers skills by keyword or intent. Returns matching skills ranked by relevance.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        query: { type: 'string', description: 'Natural language query or keywords, e.g. "deploy to production"' },
+      },
+      required: ['query'],
+    },
+    handler({ query }) {
+      return searchSkills(query).map(({ name, description, phase }) => ({ name, description, phase }));
+    },
+  },
+
+  bigpowers_get_state: {
+    description: 'Get the current bigpowers project state (active flow, epic, story, step) from specs/state.yaml.',
+    inputSchema: { type: 'object', properties: {} },
+    handler() {
+      if (!fs.existsSync(STATE_YAML)) return { error: 'specs/state.yaml not found' };
+      return { content: fs.readFileSync(STATE_YAML, 'utf8') };
+    },
+  },
+
+  bigpowers_invoke_skill: {
+    description: 'Get the full instructions for a skill in context. Returns SKILL.md content suitable for agent invocation.',
+    inputSchema: {
+      type: 'object',
+      properties: {
+        name: { type: 'string', description: 'Skill name to invoke (kebab-case)' },
+        context: { type: 'string', description: 'Optional context or arguments to pass with the skill invocation' },
+      },
+      required: ['name'],
+    },
+    handler({ name, context }) {
+      const content = getSkillContent(name);
+      if (!content) return { error: `Skill "${name}" not found` };
+      const result = { name, instructions: content };
+      if (context) result.context = context;
+      return result;
+    },
+  },
+};
+
+// ─── MCP stdio transport ──────────────────────────────────────────────────────
+
+let inputBuffer = Buffer.alloc(0);
+
+process.stdin.on('data', chunk => {
+  inputBuffer = Buffer.concat([inputBuffer, chunk]);
+  processBuffer();
+});
+
+process.stdin.on('end', () => process.exit(0));
+
+function processBuffer() {
+  while (true) {
+    const headerEnd = inputBuffer.indexOf('\r\n\r\n');
+    if (headerEnd === -1) break;
+
+    const header = inputBuffer.slice(0, headerEnd).toString();
+    const match = header.match(/Content-Length:\s*(\d+)/i);
+    if (!match) { inputBuffer = inputBuffer.slice(headerEnd + 4); continue; }
+
+    const contentLength = parseInt(match[1], 10);
+    const bodyStart = headerEnd + 4;
+    if (inputBuffer.length < bodyStart + contentLength) break;
+
+    const body = inputBuffer.slice(bodyStart, bodyStart + contentLength).toString();
+    inputBuffer = inputBuffer.slice(bodyStart + contentLength);
+
+    try { handleMessage(JSON.parse(body)); } catch (_) {}
+  }
+}
+
+function send(msg) {
+  const body = JSON.stringify(msg);
+  process.stdout.write(`Content-Length: ${Buffer.byteLength(body, 'utf8')}\r\n\r\n${body}`);
+}
+
+function handleInitialize(id) {
+  send({
+    jsonrpc: '2.0', id,
+    result: {
+      protocolVersion: '2024-11-05',
+      capabilities: { tools: {} },
+      serverInfo: { name: 'bigpowers', version: '1.0.0' },
+    },
+  });
+}
+
+function handleToolsList(id) {
+  send({
+    jsonrpc: '2.0', id,
+    result: {
+      tools: Object.entries(TOOLS).map(([name, tool]) => ({
+        name,
+        description: tool.description,
+        inputSchema: tool.inputSchema,
+      })),
+    },
+  });
+}
+
+function handleToolCall(id, params) {
+  const { name, arguments: args } = params;
+  const tool = TOOLS[name];
+  if (!tool) {
+    send({ jsonrpc: '2.0', id, error: { code: -32601, message: `Unknown tool: ${name}` } });
+    return;
+  }
+  try {
+    const result = tool.handler(args || {});
+    send({
+      jsonrpc: '2.0', id,
+      result: { content: [{ type: 'text', text: JSON.stringify(result, null, 2) }] },
+    });
+  } catch (err) {
+    send({ jsonrpc: '2.0', id, error: { code: -32603, message: err.message } });
+  }
+}
+
+function handleMessage(msg) {
+  const { id, method, params } = msg;
+  if (method === 'initialize') { handleInitialize(id); return; }
+  if (method === 'notifications/initialized') return;
+  if (method === 'tools/list') { handleToolsList(id); return; }
+  if (method === 'tools/call') { handleToolCall(id, params); return; }
+  if (id !== undefined) {
+    send({ jsonrpc: '2.0', id, error: { code: -32601, message: `Method not found: ${method}` } });
+  }
+}
+
+// Allow require() without starting the server (for testing)
+if (require.main === module) {
+  process.stderr.write('bigpowers MCP server started (stdio)\n');
+}

package/skills-lock.json

@@ -18,12 +18,12 @@
     },
     "build-epic": {
       "description": "Eight-step epic build cycle — reads state.yaml, execution-status.yaml, and one epic capsule; updates status via bp-yaml-set or direct edit. Resume mode runs one step per invocation. Use instead of ad-hoc execute-plan for release work.",
-      "sha256": "3c1707cca32b24de",
+      "sha256": "b7049084cd180053",
       "path": "build-epic/SKILL.md"
     },
     "change-request": {
       "description": "Add a new requirement or reorder epics by WSJF against specs/release-plan.yaml and epic capsule directories. Modes Add and Reorder. Use when a new requirement arrives mid-release or the plan needs prioritization.",
-      "sha256": "431cfd2a21ac4515",
+      "sha256": "1d1d7a032226b924",
       "path": "change-request/SKILL.md"
     },
     "commit-message": {
@@ -203,7 +203,7 @@
     },
     "release-branch": {
       "description": "Make the merge/PR/keep/discard decision for a feature branch, verify coverage gates, create the PR with gh, and clean up the worktree. Use when a feature is done and ready to ship, or when user says \"release\", \"merge\", or \"open a PR\".",
-      "sha256": "140f4bec3690bd51",
+      "sha256": "5d1e07e58d1aff44",
       "path": "release-branch/SKILL.md"
     },
     "request-review": {