feat(v3.3.0): wiki knowledge base + auto-research loop + multi-LLM council

README.md

@@ -13,8 +13,8 @@
 <h3 align="center">Your Claude Code gets smarter every session.</h3>
 
 <p align="center">
-  Self-correcting memory that compounds over 50+ sessions. You correct Claude once &mdash; it never makes the same mistake again.<br/>
-  <b>24 skills</b> &bull; <b>8 agents</b> &bull; <b>21 commands</b> &bull; <b>29 hook scripts across 24 events</b><br/>
+  Self-correcting memory that compounds over 50+ sessions. You correct Claude once &mdash; it never makes the same mistake again. Persistent research wikis indexed in FTS5 surface relevant prior work the moment you ask. Auto-research loop grows the knowledge base while you sleep.<br/>
+  <b>33 skills</b> &bull; <b>8 agents</b> &bull; <b>22 commands</b> &bull; <b>37 hook scripts across 24 events</b><br/>
   Works with <b>Claude Code</b>, <b>Cursor</b>, and <b>32+ agents</b> via SkillKit.
 </p>
 
@@ -75,6 +75,50 @@ cd ~/.claude/plugins/*/pro-workflow && npm install && npm run build
 
 ---
 
+## Wiki Knowledge Base — 60-second tour
+
+Auto-grow a persistent FTS5-indexed research wiki next to your code:
+
+```bash
+# Scaffold
+/wiki init agent-memory --title "Agent Memory" --flavor research
+
+# Add a page (any .md you wrote — file gets FTS-indexed)
+/wiki page agent-memory wiki/concepts/episodic-memory.md --type concept
+
+# Ask the wiki anything
+/wiki ask "what is episodic memory" --wiki agent-memory
+
+# Auto-grow it: queue a seed, run the loop, watch it write pages + enqueue follow-ups
+/wiki seed agent-memory "memory consolidation in agents"
+/wiki research agent-memory --max-pages 5 --budget-usd 0.50
+
+# Hybrid retrieval (BM25 + vector RRF) — needs OPENAI_API_KEY or VOYAGE_API_KEY
+/wiki embed agent-memory
+/wiki hybrid "consolidation patterns" --wiki agent-memory
+
+# Multi-LLM deliberation, transcript persisted as a wiki page
+/wiki council "should we adopt episodic memory?" --wiki agent-memory
+```
+
+Every wiki hit auto-loads on `UserPromptSubmit` when prompts mention indexed topics. Kill-switch: `touch ~/.pro-workflow/STOP`.
+
+---
+
+## What's New in v3.3
+
+Persistent knowledge plane on top of the self-correction memory.
+
+- **Wiki Builder** &mdash; Persistent research wikis on disk + SQLite FTS5 shadow index. 9 flavors: research, paper, domain, product, person, organization, project, codebase, incident. `/wiki init`, `/wiki page`, `/wiki reindex`.
+- **Wiki Query** &mdash; BM25 retrieval across wiki pages. `/wiki ask "<query>"` returns top-K with citations. `related` and `show` subcommands. Auto-loads top-3 hits on UserPromptSubmit when prompt mentions indexed topics.
+- **Wiki-scoped learnings** &mdash; `[LEARN] ... Wiki: <slug>` binds a rule to one wiki, no cross-project pollution.
+- **Auto-research loop** &mdash; Budget-capped BFS driver. Pluggable source fetchers (web/arXiv/GitHub + custom). Convergence detection, kill-switch, depth caps. `/wiki seed`, `/wiki research`, `/wiki seeds`, `/wiki cancel`, `/wiki status`.
+- **Hybrid retrieval (optional)** &mdash; sqlite-vec compatible embeddings via OpenAI or Voyage. RRF fusion of BM25 + vector. `/wiki embed`, `/wiki hybrid`. Degrades cleanly to BM25-only when no embedding key set.
+- **LLM Council** &mdash; Provider-agnostic 3-phase deliberation (Anthropic/OpenAI/OpenRouter/Fireworks/custom OpenAI-compat). Persists transcript as a wiki page when `--wiki` is passed. `/wiki council`.
+- **Survey Generator** &mdash; Provider-agnostic literature survey artifact. Output target = wiki markdown page (not standalone HTML). Bibliography rows append to `sources.md` deduped. `/wiki survey`.
+- **Reactive triggers** &mdash; Edits inside a wiki tree auto-enqueue verify-seeds. Cron-tick script (`scripts/research-tick.js`) runs one iteration of the oldest opted-in wiki with pending seeds.
+- **Schema additions** &mdash; `wikis`, `wiki_pages` (+ FTS5), `wiki_sources`, `wiki_claims`, `wiki_seeds`, `wiki_embeddings`, `learnings_wiki`.
+
 ## What's New in v3.2
 
 - **LLM Gates** &mdash; First plugin with `type: "prompt"` hooks for AI-powered commit validation and secret detection
@@ -94,15 +138,19 @@ cd ~/.claude/plugins/*/pro-workflow && npm install && npm run build
 | Feature | Pro Workflow | [Superpowers](https://github.com/obra/superpowers) | [ECC](https://github.com/affaan-m/everything-claude-code) | [gstack](https://github.com/garrytan/gstack) | [GSD](https://github.com/gsd-build/get-shit-done) |
 |---------|:-----------:|:-----------:|:---:|:------:|:---:|
 | Self-correcting memory (SQLite + FTS5) | **Yes** | No | No | No | No |
+| Persistent research wikis (FTS5-indexed) | **Yes** | No | No | No | No |
+| Auto-research loop (budget-capped BFS) | **Yes** | No | No | No | No |
+| Hybrid retrieval (BM25 + vector + RRF) | **Yes** | No | No | No | No |
+| Multi-provider LLM council | **Yes** | No | No | No | No |
 | LLM-powered hooks (`type: "prompt"`) | **Yes** | No | No | No | No |
 | Permission denial analysis | **Yes** | No | No | No | No |
 | Compaction-aware state preservation | **Yes** | No | No | No | No |
 | Cost tracking and budget alerts | **Yes** | No | No | No | No |
 | MCP overhead auditing | **Yes** | No | No | No | No |
 | Cross-agent (32+ agents via SkillKit) | **Yes** | No | Some | No | No |
-| Skills | 24 | 14 | 140+ | 18+ | 0 |
+| Skills | 33 | 14 | 140+ | 18+ | 0 |
 | Agents | 8 | 5 | 36 | 0 | 18 |
-| Commands | 21 | 3 | 60+ | 5+ | 57 |
+| Commands | 22 | 3 | 60+ | 5+ | 57 |
 | Hook Events | 24 | 8 | 18 | 0 | 0 |
 
 ---

commands/doctor.md

@@ -82,6 +82,22 @@ cat ~/.claude/settings.json 2>/dev/null | head -3
 - Check for conflicting settings between project and user level
 - Verify permission rules are correct
 
+### 8. Wiki Knowledge Base
+```bash
+node $PRO_WORKFLOW_ROOT/skills/wiki-builder/scripts/wiki-cli.js list 2>/dev/null
+node $PRO_WORKFLOW_ROOT/skills/wiki-research-loop/scripts/research-loop.js status 2>/dev/null
+ls -1 ~/.pro-workflow/STOP 2>/dev/null && echo "KILL SWITCH ACTIVE — no research runs"
+tail -5 ~/.pro-workflow/tick.log 2>/dev/null
+```
+- List of wikis · seed counts per status · kill-switch state · last cron-tick activity
+- Embeddings: `OPENAI_API_KEY`/`VOYAGE_API_KEY` set? Hybrid search uses provider; otherwise BM25-only.
+
+### 9. Council Providers
+```bash
+node $PRO_WORKFLOW_ROOT/skills/llm-council/scripts/council.js providers 2>/dev/null
+```
+- Shows which provider env vars are set (Anthropic/OpenAI/OpenRouter/Fireworks/custom).
+
 ## Quick Fixes
 
 | Issue | Fix |

commands/wiki.md

@@ -0,0 +1,83 @@
+# /wiki — Persistent Research Wikis
+
+Build, query, and maintain long-lived knowledge bases. Each wiki = markdown folder + SQLite FTS5 shadow index. Survives sessions, indexes auto-load on `SessionStart`.
+
+## Subcommands
+
+| Subcommand | Action |
+|------------|--------|
+| `/wiki init <slug> --title "X" [--flavor research] [--scope global\|project]` | Scaffold + register a new wiki |
+| `/wiki list [--scope global\|project]` | List registered wikis |
+| `/wiki info <slug>` | Show wiki metadata + page count |
+| `/wiki page <slug> <rel-path> [--title "X"] [--type concept\|paper\|...]` | Upsert a markdown page into the FTS index |
+| `/wiki reindex <slug>` | Walk `wiki/` and re-index all `*.md` |
+| `/wiki ask "<query>" [--wiki <slug>] [--limit N]` | BM25 search over wiki pages |
+| `/wiki related <slug> <rel-path>` | Find adjacent pages by reusing title+summary as query |
+| `/wiki show <slug> <rel-path>` | Print a page from the index |
+| `/wiki seed <slug> "<query>" [--depth 0]` | Enqueue a research seed |
+| `/wiki research <slug> [--max-pages 5] [--budget-usd 0.50] [--fetchers web,arxiv,github]` | Run the auto-research loop |
+| `/wiki seeds <slug> [--status pending\|active\|done\|failed]` | List queued seeds |
+| `/wiki cancel <slug>` | Mark all pending/active seeds as failed |
+| `/wiki status` | Cross-wiki seed counts + kill-switch state |
+| `/wiki embed [<slug>] [--limit N]` | Compute embeddings for indexed pages (needs `OPENAI_API_KEY` or `VOYAGE_API_KEY`) |
+| `/wiki hybrid "<query>" [--wiki <slug>]` | Hybrid retrieval: BM25 + vector + RRF |
+| `/wiki council "<query>" --wiki <slug>` | Run llm-council; persist transcript as a wiki page |
+| `/wiki survey --bundle <path> --wiki <slug>` | Generate literature survey from a research_bundle.json |
+
+## Routes to skills
+
+- `init / list / info / page / reindex` → `wiki-builder`
+- `ask / related / show / hybrid` → `wiki-query` (+ `embed-wiki.js` for hybrid)
+- `seed / research / seeds / cancel / status` → `wiki-research-loop`
+- `council` → `llm-council`
+- `survey` → `survey-generator`
+- `embed` → `scripts/embed-wiki.js`
+
+## Storage
+
+- **Global** (default): `~/.pro-workflow/wikis/<slug>/`
+- **Project**: `<project>/.claude/wikis/<slug>/` via `--scope project`
+
+Both indexed in `~/.pro-workflow/data.db`.
+
+## Flavors
+
+`research`, `paper`, `domain`, `product`, `person`, `organization`, `project`, `codebase`, `incident`. See wiki-builder skill for layout per flavor.
+
+## Examples
+
+```
+/wiki init agent-memory --title "Agent Memory" --flavor research
+/wiki page agent-memory wiki/concepts/episodic-memory.md --type concept
+/wiki ask "what is episodic memory" --wiki agent-memory
+/wiki related agent-memory wiki/concepts/episodic-memory.md
+```
+
+## Auto-research (Phase 3.3.1+)
+
+Loop is opt-in per-wiki via `wiki.config.md`:
+
+```yaml
+auto_research:
+  enabled: true
+  max_pages_per_run: 5
+  max_depth: 3
+  budget_usd: 0.50
+```
+
+Phase 3.3.0 ships the index + manual page workflow. Loop driver lands next.
+
+## Cited recall
+
+Every answer that uses a wiki hit must cite as:
+
+```
+[wiki:<slug>] <Title> — `<rel_path>`
+```
+
+No paraphrase without citation.
+
+## Related
+
+- `/learn` (and `learn-rule`) accepts `--wiki <slug>` to scope a learning rule to one wiki.
+- SessionStart auto-loads top-3 wiki hits when prompt matches indexed content.

docs/infographic.html

@@ -680,7 +680,7 @@
       <div class="header-sub" style="margin-left: 18px;">github.com/rohitg00/pro-workflow &middot; <a href="https://rohitg00-pro-workflow.mintlify.app/" style="color:var(--anthro-coral); text-decoration:none;">docs</a></div>
     </div>
     <div class="header-right">
-      <div class="star-count">v3.2</div>
+      <div class="star-count">v3.3</div>
       <div class="star-label">Latest Release</div>
     </div>
   </div>
@@ -690,8 +690,8 @@
     <div class="title">Pro Workflow</div>
     <div class="subtitle">
       Battle-tested Claude Code patterns from power users who ship production code daily.
-      Self-correcting memory, parallel worktrees, quality gates, and the 80/20 AI coding ratio &mdash;
-      24 Skills, 8 Agents, 24 Hooks, 21 Commands, 3 Contexts, 7 Guides.
+      Self-correcting memory, persistent FTS5-indexed wikis, auto-research loop, multi-LLM council, parallel worktrees, quality gates &mdash;
+      33 Skills, 8 Agents, 24 Hook events, 22 Commands, 3 Contexts, 7 Guides.
     </div>
   </div>
 
@@ -751,10 +751,10 @@
       <div class="section-num">03 Components</div>
       <div class="section-title">Everything Included</div>
       <div class="comp-grid" style="grid-template-columns: repeat(4, 1fr);">
-        <div class="comp-item"><div class="comp-num">24</div><div class="comp-label">Skills</div></div>
+        <div class="comp-item"><div class="comp-num">33</div><div class="comp-label">Skills</div></div>
         <div class="comp-item"><div class="comp-num">8</div><div class="comp-label">Agents</div></div>
-        <div class="comp-item"><div class="comp-num">24</div><div class="comp-label">Hooks</div></div>
-        <div class="comp-item"><div class="comp-num">21</div><div class="comp-label">Commands</div></div>
+        <div class="comp-item"><div class="comp-num">24</div><div class="comp-label">Hook events</div></div>
+        <div class="comp-item"><div class="comp-num">22</div><div class="comp-label">Commands</div></div>
         <div class="comp-item"><div class="comp-num">3</div><div class="comp-label">Contexts</div></div>
         <div class="comp-item"><div class="comp-num">7</div><div class="comp-label">Rules</div></div>
         <div class="comp-item"><div class="comp-num">29</div><div class="comp-label">Scripts</div></div>
@@ -1161,29 +1161,34 @@
     <!-- 10 SKILLS -->
     <div>
       <div class="section-num">10 Skills Library</div>
-      <div class="section-title">24 Skills</div>
+      <div class="section-title">33 Skills</div>
       <table>
         <tr><th>Skill</th><th>What It Does</th></tr>
         <tr><td><strong>pro-workflow</strong></td><td>Core 8 patterns</td></tr>
+        <tr style="background:var(--anthro-light-coral);"><td><strong>wiki-builder</strong></td><td>Persistent FTS5-indexed research wikis (9 flavors)</td></tr>
+        <tr style="background:var(--anthro-light-coral);"><td><strong>wiki-query</strong></td><td>BM25 retrieval, snippets, related/show</td></tr>
+        <tr style="background:var(--anthro-light-coral);"><td><strong>wiki-research-loop</strong></td><td>Budget-capped BFS auto-research over web/arXiv/GH</td></tr>
+        <tr style="background:var(--anthro-light-coral);"><td><strong>llm-council</strong></td><td>Provider-agnostic 3-phase multi-LLM deliberation</td></tr>
+        <tr style="background:var(--anthro-light-coral);"><td><strong>survey-generator</strong></td><td>Provider-agnostic literature survey to wiki page</td></tr>
         <tr><td><strong>smart-commit</strong></td><td>Quality gates + conventional commits</td></tr>
         <tr><td><strong>wrap-up</strong></td><td>Session ritual with learning capture</td></tr>
-        <tr><td><strong>learn-rule</strong></td><td>Persist corrections to memory</td></tr>
+        <tr><td><strong>learn-rule</strong></td><td>Persist corrections to memory (wiki-scoped)</td></tr>
         <tr><td><strong>parallel-worktrees</strong></td><td>Git worktree setup for zero dead time</td></tr>
         <tr><td><strong>replay-learnings</strong></td><td>Surface relevant past patterns</td></tr>
         <tr><td><strong>session-handoff</strong></td><td>Resume docs for next session</td></tr>
         <tr><td><strong>insights</strong></td><td>Analytics, heatmaps, trends</td></tr>
         <tr><td><strong>deslop</strong></td><td>Remove AI code slop before commit</td></tr>
         <tr><td><strong>context-optimizer</strong></td><td>Token management and MCP audit</td></tr>
         <tr><td><strong>orchestrate</strong></td><td>Multi-phase dev workflow</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>permission-tuner</strong></td><td>Optimize allow/deny rules</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>compact-guard</strong></td><td>Protect context through compaction</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>cost-tracker</strong></td><td>Session cost awareness + budgets</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>auto-setup</strong></td><td>Auto-configure quality gates</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>mcp-audit</strong></td><td>MCP overhead analysis</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>llm-gate</strong></td><td>AI-powered quality gates</td></tr>
-        <tr style="background:var(--anthro-light-coral);"><td><strong>file-watcher</strong></td><td>Reactive file change workflows</td></tr>
+        <tr><td><strong>permission-tuner</strong></td><td>Optimize allow/deny rules</td></tr>
+        <tr><td><strong>compact-guard</strong></td><td>Protect context through compaction</td></tr>
+        <tr><td><strong>cost-tracker</strong></td><td>Session cost awareness + budgets</td></tr>
+        <tr><td><strong>auto-setup</strong></td><td>Auto-configure quality gates</td></tr>
+        <tr><td><strong>mcp-audit</strong></td><td>MCP overhead analysis</td></tr>
+        <tr><td><strong>llm-gate</strong></td><td>AI-powered quality gates</td></tr>
+        <tr><td><strong>file-watcher</strong></td><td>Reactive file change workflows</td></tr>
       </table>
-      <div style="font-size:10px; color:var(--anthro-muted); margin-top:6px;">+ 7 more: context-engineering, batch-orchestration, thoroughness-scoring, sprint-status, agent-teams, safe-mode</div>
+      <div style="font-size:10px; color:var(--anthro-muted); margin-top:6px;">Coral rows = new in v3.3 (knowledge plane). + 10 more: context-engineering, batch-orchestration, thoroughness-scoring, sprint-status, agent-teams, safe-mode, token-efficiency, anti-sycophancy, tool-call-budget, design-rules</div>
     </div>
 
     <!-- 11 MODEL SELECTION -->

package.json

@@ -1,11 +1,11 @@
 {
   "name": "pro-workflow",
-  "version": "3.2.0",
-  "description": "Complete AI coding workflow system with orchestration patterns, cross-agent support, reference guides, and searchable learnings",
+  "version": "3.3.0",
+  "description": "Complete AI coding workflow system with orchestration patterns, cross-agent support, reference guides, searchable learnings, and persistent research wikis",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
   "scripts": {
-    "build": "tsc",
+    "build": "tsc && node -e \"const fs=require('fs');fs.mkdirSync('dist/db',{recursive:true});fs.copyFileSync('src/db/schema.sql','dist/db/schema.sql')\"",
     "clean": "rm -rf dist",
     "prepublishOnly": "npm run build",
     "db:init": "node dist/db/index.js"