wicked-brain 0.17.2 → 0.18.1

package/README.md

@@ -10,7 +10,22 @@ __      _(_) ___| | _____  __| |     | |__  _ __ __ _(_)_ __
 
 wicked-brain gives your AI coding CLI a persistent, searchable knowledge base built on markdown and SQLite. Drop in files, let the agent organize them, and query your accumulated knowledge across sessions — all without leaving your terminal.
 
-Works with **Claude Code**, **Gemini CLI**, **Copilot CLI**, **Cursor**, and **Codex**.
+Works with **Claude Code**, **Gemini CLI**, **Copilot CLI**, **Cursor**, **Codex**, **Kiro**, and **Antigravity**.
+
+> **Status:** v0.17.2, published to npm as [`wicked-brain`](https://www.npmjs.com/package/wicked-brain).
+> Actively developed (JS). This is a **bridge-period adapter** — its memory and knowledge role is
+> destined to fold into [wicked-estate](https://github.com/mikeparcewski/wicked-estate).
+>
+> **The differentiator:** a RAG alternative that stores knowledge as human-readable markdown with
+> explicit, confidence-scored `[[backlinks]]` instead of opaque embeddings — every claim traces to a
+> source file, and the agent (not a vector index) does the reasoning.
+
+wicked-brain is the **bridge-period memory** of the [wicked-* foundation](https://we.wickedagile.com):
+a local-first stack for AI coding agents anchored by
+[wicked-estate](https://github.com/mikeparcewski/wicked-estate) (the code graph), with
+[wicked-core](https://github.com/mikeparcewski/wicked-core) (the runtime),
+[wicked-bus](https://github.com/mikeparcewski/wicked-bus) (the durable event fabric), and
+[wicked-crew](https://github.com/mikeparcewski/wicked-crew) (the agentic execution harness).
 
 ---
 
@@ -161,6 +176,10 @@ The viewer has no auth. It's localhost-only, same-machine trust.
 | `wicked-brain:lsp` | Universal code intelligence via LSP — hover, go-to-definition, diagnostics, completions |
 | `wicked-brain:graph` | Code-relationship graph — blast radius, callers, lineage — backed by a static code graph |
 | `wicked-brain:ui` | Open the read-only browser viewer — Material-styled Search + Wiki tabs over `http://localhost:<port>/` |
+| `wicked-brain:context` | Surface relevant brain knowledge for the current prompt — runs inline on the hot path to enrich what you're working on |
+| `wicked-brain:onboard` | Full project-understanding pipeline — scans the repo, investigates from multiple perspectives, and builds the support wiki |
+| `wicked-brain:session-teardown` | Capture session learnings — decisions, patterns, gotchas, discoveries — as brain memories before a session ends |
+| `wicked-brain:consolidate` | Multi-pass brain maintenance — archive noise, promote patterns, merge duplicates, and rebuild the synonym map |
 
 ## Code Graph (offline / air-gapped)
 
@@ -279,7 +298,7 @@ Modern LLMs read PDF, DOCX, PPTX, and XLSX natively. When you ingest a binary do
 
 ## Architecture
 
-Plain Node.js server (SQLite FTS5 + file watcher + optional LSP client + HTML viewer) plus markdown skill instructions your AI CLI consumes. **One runtime dependency** (`better-sqlite3`); LSP layer is hand-rolled JSON-RPC; the viewer is vanilla JS with no build.
+Plain Node.js server (SQLite FTS5 + file watcher + optional LSP client + HTML viewer) plus markdown skill instructions your AI CLI consumes. **Two runtime dependencies** (`better-sqlite3` and `wicked-bus`); LSP layer is hand-rolled JSON-RPC; the viewer is vanilla JS with no build.
 
 Compare that to a typical RAG stack:
 
@@ -293,7 +312,7 @@ Typical RAG:                           wicked-brain:
 - Orchestration layer                 - Skills (markdown)
 - Admin UI                            - GET / (vanilla HTML, read-only)
 ─────────────────                     ─────────────────
-10+ deps, opaque vectors              1 runtime dep, plain markdown
+10+ deps, opaque vectors              2 runtime deps, plain markdown
 ```
 
 See [ARCHITECTURE.md](ARCHITECTURE.md) for component diagrams and the schema layout.

package/install.mjs

@@ -1,5 +1,6 @@
 #!/usr/bin/env node
-// wicked-brain installer — detects CLIs and installs skills + agents
+// wicked-brain installer — detects CLIs and installs skills (skills-only;
+// no per-platform agent materialization).
 
 import { existsSync, mkdirSync, cpSync, readdirSync } from "node:fs";
 import { join, resolve, basename } from "node:path";
@@ -21,8 +22,6 @@ function buildClaudeTarget(rootDir, source, { trusted = false } = {}) {
     name: "claude",
     rootDir,
     dir: join(rootDir, "skills"),
-    agentDir: join(rootDir, "agents"),
-    agentSubdir: "agents",
     platform: "claude",
     identityMarkers: ["settings.json", "plugins", "projects"],
     source,
@@ -49,12 +48,16 @@ function resolveClaudeCandidates() {
 // Canonical non-claude targets. Claude is expanded dynamically via
 // resolveClaudeCandidates() below so CLI_TARGETS stays a flat spec.
 const CLI_TARGETS = [
-  { name: "gemini",      dir: join(home, ".gemini", "skills"),      agentDir: join(home, ".gemini", "agents"),      agentSubdir: "agents", platform: "gemini" },
-  { name: "copilot",     dir: join(home, ".github", "skills"),      agentDir: join(home, ".github", "agents"),      agentSubdir: "agents", platform: "copilot" },
-  { name: "codex",       dir: join(home, ".codex", "skills"),       agentDir: join(home, ".codex", "agents"),       agentSubdir: "agents", platform: "codex" },
-  { name: "cursor",      dir: join(home, ".cursor", "skills"),      agentDir: join(home, ".cursor", "agents"),      agentSubdir: "agents", platform: "cursor" },
-  { name: "kiro",        dir: join(home, ".kiro", "skills"),        agentDir: join(home, ".kiro", "agents"),        agentSubdir: "agents", platform: "kiro" },
-  { name: "antigravity", dir: join(home, ".antigravity", "skills"), agentDir: join(home, ".antigravity", "rules"),  agentSubdir: "rules",  platform: "antigravity" },
+  // Paths verified against each CLI's own source of truth:
+  //  - Copilot CLI loads personal custom skills from ~/.copilot/skills/ (not ~/.github/skills).
+  //  - Antigravity / Gemini CLI load global custom skills from ~/.gemini/config/skills/
+  //    (confirmed by the Antigravity agent; not ~/.gemini/skills or ~/.antigravity/skills).
+  { name: "gemini",      dir: join(home, ".gemini", "config", "skills"), platform: "gemini" },
+  { name: "copilot",     dir: join(home, ".copilot", "skills"),          platform: "copilot" },
+  { name: "codex",       dir: join(home, ".codex", "skills"),            platform: "codex" },
+  { name: "cursor",      dir: join(home, ".cursor", "skills"),           platform: "cursor" },
+  { name: "kiro",        dir: join(home, ".kiro", "skills"),             platform: "kiro" },
+  { name: "antigravity", dir: join(home, ".gemini", "config", "skills"), platform: "antigravity" },
 ];
 
 // Identity-marker gate for claude candidates. Without this, probing
@@ -112,14 +115,13 @@ let targets;
 
 if (pathArg && typeof pathArg === "string" && pathArg !== "") {
   const customPath = resolve(pathArg.replace(/^~/, () => home));
-  // Strip leading dot to match CLI_TARGETS names (e.g. ".claude" → "claude")
+  // Strip leading dot to match CLI_TARGETS names (e.g. ".claude" → "claude").
+  // The matched platform is still used to select the right hook template.
   const dirName = basename(customPath).replace(/^\./, "");
   const knownPlatform = CLI_TARGETS.find((t) => t.name === dirName);
-  const agentSubdir = knownPlatform?.agentSubdir ?? "agents";
   targets = [{
     name: dirName,
     dir: join(customPath, "skills"),
-    agentDir: join(customPath, agentSubdir),
     platform: knownPlatform?.platform ?? dirName,
   }];
   console.log(`Custom path: ${customPath}\n`);
@@ -151,6 +153,20 @@ if (pathArg && typeof pathArg === "string" && pathArg !== "") {
 
   const cliFilter = (typeof cliArg === "string" && cliArg !== "") ? cliArg.split(",") : null;
   targets = cliFilter ? detected.filter((d) => cliFilter.includes(d.name)) : detected;
+
+  // A non-empty --cli filter that matches none of the detected CLIs must be a
+  // hard error. Otherwise targets is [], we copy zero skills, and exit 0 — a
+  // silent false success that looks like a clean install. Name the unmatched
+  // value(s) and what was actually detected so the user can correct the flag.
+  if (cliFilter && targets.length === 0) {
+    const detectedNames = detected.map((d) => d.name);
+    const unmatched = cliFilter.filter((n) => !detectedNames.includes(n));
+    console.error(
+      `Error: --cli filter matched no detected CLI. Unmatched: ${unmatched.join(", ")}. ` +
+      `Detected: ${detectedNames.join(", ")}.`
+    );
+    process.exit(1);
+  }
 }
 
 // Copy skills to each target CLI
@@ -169,31 +185,10 @@ for (const target of targets) {
   console.log(`  ${skillDirs.length} skills installed`);
 }
 
-// Copy platform-specific agents
-const agentsSource = join(__dirname, "skills", "wicked-brain-agent", "platform");
-
-for (const target of targets) {
-  const platformDir = join(agentsSource, target.platform);
-  if (!existsSync(platformDir)) {
-    console.log(`  No agent definitions for ${target.name}, skipping agents`);
-    continue;
-  }
-
-  if (!target.agentDir) continue;
-
-  mkdirSync(target.agentDir, { recursive: true });
-  const agentFiles = readdirSync(platformDir);
-  let agentCount = 0;
-
-  for (const file of agentFiles) {
-    const src = join(platformDir, file);
-    const dest = join(target.agentDir, file);
-    cpSync(src, dest, { force: true });
-    agentCount++;
-  }
-
-  console.log(`  Installed ${agentCount} agents to ${target.agentDir}`);
-}
+// Skills-only distribution: behavior that used to be materialized into
+// per-platform native agent files now lives entirely in first-class skills
+// (wicked-brain-consolidate / -onboard / -session-teardown / -context),
+// copied above like any other skill. No agent materialization step remains.
 
 // Optional hook installation (--hooks flag). Goes through flagValue so
 // `--hooks=false` correctly disables; bare `--hooks` and `--hooks=true`
@@ -202,7 +197,7 @@ const installHooks = flagValue("hooks") === true;
 
 if (installHooks) {
   console.log("\nInstalling hooks...");
-  const hooksSource = join(__dirname, "skills", "wicked-brain-agent", "hooks");
+  const hooksSource = join(__dirname, "skills", "wicked-brain-context", "hooks");
 
   for (const target of targets) {
     const hookFile = join(hooksSource, `${target.platform}-hooks.json`);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "wicked-brain",
-  "version": "0.17.2",
+  "version": "0.18.1",
   "type": "module",
   "description": "Digital brain as skills for AI coding CLIs — no vector DB, no embeddings, no infrastructure",
   "keywords": [

package/server/package.json

@@ -1,6 +1,6 @@
 {
   "name": "wicked-brain-server",
-  "version": "0.17.2",
+  "version": "0.18.1",
   "type": "module",
   "description": "SQLite FTS5 search server for wicked-brain digital knowledge bases",
   "keywords": [

package/skills/wicked-brain-configure/SKILL.md

@@ -38,7 +38,7 @@ Read frontmatter of top results, count `contains:` tag frequency. Top 10 tags =
 
 4. Read `{brain_path}/_meta/log.jsonl` (last 50 lines) for recent `search_miss` entries — these are knowledge gaps.
 
-5. List available agents by reading `skills/wicked-brain-agent/agents/` at depth 0.
+5. Note the brain workflow skills available: `wicked-brain:context` (surface context), `wicked-brain:session-teardown` (capture learnings), `wicked-brain:consolidate` (maintenance), and `wicked-brain:onboard` (project understanding).
 
 ### Step 2: Detect CLI/IDE
 
@@ -85,10 +85,10 @@ Digital brain: {brain_id} | {total} indexed items | {chunks} chunks, {wiki} wiki
 - **Search/explore**: use `wicked-brain:search` — replaces Grep, Glob, and Agent(Explore) for any open-ended search
 - **Answer questions**: use `wicked-brain:query` — replaces Agent(Explore) for conceptual questions
 - **Wiki catalog**: use `wicked-brain:read` at depth 0/1 to browse wiki articles progressively
-- **Surface context**: call `wicked-brain:agent` (context) at the start of any new topic
-- **Capture learnings**: call `wicked-brain:agent` (session-teardown) at session end
+- **Surface context**: call `wicked-brain:context` at the start of any new topic
+- **Capture learnings**: call `wicked-brain:session-teardown` at session end
 - **Store a decision/pattern/gotcha**: call `wicked-brain:memory` (store mode)
-- **Available agents**: consolidate, context, session-teardown, onboard (via `wicked-brain:agent`)
+- **Workflow skills**: `wicked-brain:consolidate`, `wicked-brain:context`, `wicked-brain:session-teardown`, `wicked-brain:onboard`
 
 ### Search result source types
 
@@ -104,9 +104,9 @@ Brain search/query results include `source_type` and `path` fields:
 - The server auto-starts on any `wicked-brain-call` / skill invocation — a "server not running" notice is NEVER a reason to skip the brain; just call the skill
 - Use `wicked-brain:search` for finding content ("find X", "where is Y", "look for Z", "show me W")
 - Use `wicked-brain:query` for questions ("what does X do", "how does Y work", "explain Z")
-- Use `wicked-brain:agent` (context) when starting a new topic or unfamiliar area
+- Use `wicked-brain:context` when starting a new topic or unfamiliar area
 - Only fall back to Grep/Glob for **exact pattern matching** after the brain returns no results
-- Do NOT read brain files directly — always go through skills and agents
+- Do NOT read brain files directly — always go through skills
 - Always pass `session_id` with search/query calls for access tracking
 - Capture non-obvious decisions, patterns, and gotchas with `wicked-brain:memory`
 - When search results include `source_type: wiki`, follow up with `wicked-brain:read` at depth 1-2

package/skills/{wicked-brain-agent/agents/consolidate.md → wicked-brain-consolidate/SKILL.md}

@@ -1,19 +1,61 @@
-# consolidate
+---
+name: wicked-brain:consolidate
+description: |
+  Multi-pass brain consolidation - archive noise, promote patterns, merge
+  duplicates, build synonym map.
 
-## Depth 0 — Summary
-Four-pass lifecycle: archive noise, promote patterns, merge duplicates, build synonym map.
-Triggers: manual invocation via `wicked-brain:agent dispatch consolidate`, or after reviewing consolidation_hint entries in log.jsonl.
+  Use when: brain needs maintenance, after significant ingestion, or reviewing
+  consolidation_hint entries in log.jsonl.
+model: sonnet
+allowed-tools: Read, Write, Edit, Bash, Grep, Glob
+context: fork
+---
 
-## Depth 1 — Pipeline Steps
-Pass 1 (Archive): Call server candidates(archive) → review at depth 0 → check TTL expiry for memories → archive stale items + remove from index
-Pass 2 (Promote): Call server candidates(promote) → for chunks: log promote_candidate → for memories: update tier + bump confidence
-Pass 3 (Merge): Read promote candidates at depth 2 → LLM similarity comparison → keep highest-scored / archive duplicates / flag complementary for review
-Pass 4 (Synonyms): Aggregate synonym_hit/synonym_miss events from log → build/update learned synonym map at _meta/synonyms.json
+# wicked-brain:consolidate
+
+You are a consolidation agent for the digital brain. This is a heavy,
+multi-pass maintenance pipeline that runs in an isolated (forked) context so it
+has a longer token budget and file-writing tools for large-scale work.
+
+## Overview (pipeline)
+
+Four-pass lifecycle: archive noise, promote patterns, merge duplicates, build
+synonym map.
+
+- **Pass 1 (Archive)**: Call server candidates(archive) → review at depth 0 → check TTL expiry for memories → archive stale items + remove from index
+- **Pass 2 (Promote)**: Call server candidates(promote) → for chunks: log promote_candidate → for memories: update tier + bump confidence
+- **Pass 3 (Merge)**: Read promote candidates at depth 2 → LLM similarity comparison → keep highest-scored / archive duplicates / flag complementary for review
+- **Pass 4 (Synonyms)**: Aggregate synonym_hit/synonym_miss events from log → build/update learned synonym map at _meta/synonyms.json
 
 Parameters: brain_path, port, session_id
 Depends on: server candidates action, memory frontmatter schema, wicked-brain:memory skill
 
-## Depth 2 — Full Subagent Instructions
+## Config
+
+Resolve the brain config via the shared resolution in
+wicked-brain:init § "Resolving the brain config". In short: try
+`~/.wicked-brain/projects/{cwd_basename}/_meta/config.json` first, fall back
+to `~/.wicked-brain/_meta/config.json` (legacy flat), else trigger
+wicked-brain:init. Read the resolved file for brain path and server port.
+
+Do NOT read a bare relative `_meta/config.json` — the model will resolve it
+against the current working directory and brain files will end up in the
+project root.
+
+## Bus event (start)
+
+At the start of the run, emit a dispatch event (fire-and-forget — if the bus is
+not installed, silently skip):
+
+```bash
+npx wicked-bus emit \
+  --type "wicked.agent.dispatched" \
+  --domain "wicked-brain" \
+  --subdomain "brain.agent" \
+  --payload '{"agent":"consolidate","brain_id":"{brain_id}"}' 2>/dev/null || true
+```
+
+## Pipeline
 
 You are a consolidation agent for the digital brain at {brain_path}.
 Server: http://localhost:{port}/api
@@ -136,3 +178,25 @@ After all four passes, report:
 - Merged: {N} near-duplicates
 - Flagged for review: {N} complementary pairs
 - Synonyms: {N} added, {N} removed, {N} total in map
+
+## Bus event (completion)
+
+On completion, emit the consolidation-complete event (fire-and-forget — if the
+bus is not installed, silently skip):
+
+```bash
+npx wicked-bus emit \
+  --type "wicked.brain.consolidated" \
+  --domain "wicked-brain" \
+  --subdomain "brain" \
+  --payload '{"brain_id":"{brain_id}","archived":{N},"promoted":{M},"merged":{P}}' 2>/dev/null || true
+```
+
+## Cross-Platform Notes
+
+- `curl` is cross-platform (Windows 10+) — OK for server API calls.
+- The `mv ... .archived-$(date +%s)` rename uses shell; on Windows use the
+  equivalent (`Move-Item` / `ren`) or the agent-native move with a
+  `.archived-{unix_seconds}` suffix. Preserve the `.archived-{timestamp}`
+  convention either way.
+- All log/JSON appends should use forward-slash paths.

package/skills/{wicked-brain-agent/agents/context.md → wicked-brain-context/SKILL.md}

@@ -1,18 +1,71 @@
-# context
+---
+name: wicked-brain:context
+description: |
+  Surface relevant brain knowledge for the current prompt. Tiered routing - hot
+  path for simple prompts, fast path for complex.
 
-## Depth 0 — Summary
-Tiered knowledge surfacing for ambient context. Hot path for simple prompts (recent memories + high-confidence wiki). Fast path for complex prompts (full search + scoring pipeline).
+  Use when: starting a new topic/unfamiliar area, at the start of a work
+  session, prompt would benefit from prior decisions/patterns/wiki.
+model: haiku
+allowed-tools: Read, Bash, Grep, Glob
+---
+
+# wicked-brain:context
+
+You are a context assembly agent for the digital brain. This is a HOT-PATH
+enrichment skill: it runs INLINE in the current turn (it is what the
+UserPromptSubmit / BeforeAgent hook nudges) and must stay fast. It does NOT run
+in a forked context — return pointers to the host agent in the same turn.
+
+It emits a `wicked.agent.dispatched` bus event (see "Bus event" below) —
+fire-and-forget and non-blocking, so the hot path never waits on it. Only reads
+and returns pointers — no Write/Edit.
+
+## Overview (pipeline)
+
+Tiered knowledge surfacing for ambient context. Hot path for simple prompts
+(recent memories + high-confidence wiki). Fast path for complex prompts (full
+search + scoring pipeline).
 
-## Depth 1 — Pipeline Steps
 1. Classify prompt complexity (short/single-topic → hot, complex/multi-topic → fast)
-2. Hot path: search recent memories (7 days) + wiki (confidence > 0.8), return depth 0
+2. Hot path: recent memories (7 days) + wiki (confidence > 0.8), return depth 0
 3. Fast path: decompose query → synonym-expand → search all content → score by keyword overlap + type + tier + recency → return depth 0
 4. Agent reads deeper (depth 1/2) on promising results as needed
 
 Parameters: brain_path, port, session_id, prompt (the user's current prompt text)
 Depends on: server search action, synonym expansion
 
-## Depth 2 — Full Subagent Instructions
+## Config
+
+Resolve the brain config via the shared resolution in
+wicked-brain:init § "Resolving the brain config". In short: try
+`~/.wicked-brain/projects/{cwd_basename}/_meta/config.json` first, fall back
+to `~/.wicked-brain/_meta/config.json` (legacy flat), else trigger
+wicked-brain:init. Read the resolved file for brain path and server port.
+
+Do NOT read a bare relative `_meta/config.json` — the model will resolve it
+against the current working directory and brain files will end up in the
+project root.
+
+## Bus event
+
+At the start of the run, emit a dispatch event (fire-and-forget — if the bus is
+not installed, silently skip). Same type, domain, and subdomain as the sibling
+skills (`wicked-brain-consolidate`, `wicked-brain-onboard`,
+`wicked-brain-session-teardown`), with an `agent:context` payload.
+
+Because this is the inline hot path, run the emit non-blocking (background it)
+so returning pointers to the host agent never waits on it:
+
+```bash
+npx wicked-bus emit \
+  --type "wicked.agent.dispatched" \
+  --domain "wicked-brain" \
+  --subdomain "brain.agent" \
+  --payload '{"agent":"context","brain_id":"{brain_id}"}' 2>/dev/null || true
+```
+
+## Pipeline
 
 You are a context assembly agent for the digital brain at {brain_path}.
 Server: http://localhost:{port}/api
@@ -86,3 +139,8 @@ Context (fast path, {N} results):
 - Do NOT inject context silently — return it to the host agent for decision
 - Do NOT run both paths — pick one based on Step 1 classification
 - Do NOT spend more than 5 search calls on the hot path
+
+## Cross-Platform Notes
+
+- `curl` is cross-platform (Windows 10+) — OK for server API calls.
+- This skill only reads (Read, Bash, Grep, Glob) — no file writes.

package/skills/wicked-brain-forget/SKILL.md

@@ -12,7 +12,7 @@ description: |
 # wicked-brain:forget
 
 Archive or hard-delete a memory by id or path. Wraps the server `remove` action
-and the archive-rename convention used by wicked-brain:agent dispatch consolidate.
+and the archive-rename convention used by wicked-brain:consolidate.
 
 ## Cross-Platform Notes
 

package/skills/{wicked-brain-agent/agents/onboard.md → wicked-brain-onboard/SKILL.md}

@@ -1,9 +1,29 @@
-# onboard
+---
+name: wicked-brain:onboard
+description: |
+  Full project understanding - detect repo mode, scan, investigate 5
+  perspectives, extract symbols, ingest chunks, compile support wiki,
+  configure CLI.
+
+  Use when: onboarding a new project, re-onboarding after big changes, brain
+  has no project wiki yet.
+model: sonnet
+allowed-tools: Read, Write, Edit, Bash, Grep, Glob
+context: fork
+---
+
+# wicked-brain:onboard
+
+You are an onboarding agent for the digital brain. This is a heavy
+project-understanding pipeline that runs in an isolated (forked) context so it
+has a longer token budget and file-writing tools for large-scale compilation.
+
+## Overview (pipeline)
 
-## Depth 0 — Summary
-Full project understanding pipeline. Scans project, extracts findings from 5 perspectives (product, engineering, quality, ops, data), ingests as structured chunks, compiles a progressive-loading support wiki, and configures the CLI.
+Full project understanding pipeline. Scans project, extracts findings from 5
+perspectives (product, engineering, quality, ops, data), ingests as structured
+chunks, compiles a progressive-loading support wiki, and configures the CLI.
 
-## Depth 1 — Pipeline Steps
 0. Detect: run `wicked-brain-onboard-wiki` to classify repo mode, write `.wicked-brain/mode.json`, and stamp the contributor-wiki pointer into CLAUDE.md / AGENTS.md if present
 1. Scan: directory structure, key files, languages, frameworks, dependencies
 2. Investigate: gather facts from each of the 5 perspectives
@@ -15,13 +35,40 @@ Full project understanding pipeline. Scans project, extracts findings from 5 per
 Parameters: brain_path, port, project_path (defaults to cwd)
 Depends on: wicked-brain:ingest, wicked-brain:compile, wicked-brain:configure
 
-## Depth 2 — Full Subagent Instructions
+## Config
+
+Resolve the brain config via the shared resolution in
+wicked-brain:init § "Resolving the brain config". In short: try
+`~/.wicked-brain/projects/{cwd_basename}/_meta/config.json` first, fall back
+to `~/.wicked-brain/_meta/config.json` (legacy flat), else trigger
+wicked-brain:init. Read the resolved file for brain path and server port.
+
+Do NOT read a bare relative `_meta/config.json` — the model will resolve it
+against the current working directory and brain files will end up in the
+project root.
+
+## Bus event (start)
+
+At the start of the run, emit a dispatch event (fire-and-forget — if the bus is
+not installed, silently skip):
+
+```bash
+npx wicked-bus emit \
+  --type "wicked.agent.dispatched" \
+  --domain "wicked-brain" \
+  --subdomain "brain.agent" \
+  --payload '{"agent":"onboard","brain_id":"{brain_id}"}' 2>/dev/null || true
+```
+
+## Pipeline
 
 You are an onboarding agent for the digital brain at {brain_path}.
 Server: http://localhost:{port}/api
 Project: {project_path}
 
-Your job: deeply understand a project from 5 perspectives and produce a support wiki that serves engineers, testers, ops, and product owners — all through progressive loading so only what's needed gets loaded.
+Your job: deeply understand a project from 5 perspectives and produce a support
+wiki that serves engineers, testers, ops, and product owners — all through
+progressive loading so only what's needed gets loaded.
 
 ### Step 0: Detect repo mode and stamp wiki pointer
 
@@ -80,9 +127,19 @@ Gather facts for each perspective. You'll write these as chunks in Step 4.
 - Module map: which file owns what responsibility
 - Data flow: request lifecycle from entry to storage to response
 - Extension points: where to add new functionality (new action, new migration, new skill)
+- Conventions (so new work matches the repo's style):
+  - **Naming**: file, function, and variable naming patterns
+  - **Code style**: formatting, comment conventions, and import
+    ordering/grouping — detect from config files (`.eslintrc*`, `.eslintrc.json`,
+    `.prettierrc`, `pyproject.toml` ruff/black config, `rustfmt.toml`) and from
+    grouping in existing source
+  - **Build/Deploy/CI-CD**: build commands, deploy scripts, and CI/CD patterns
+    (`.github/workflows/`, Makefile targets, package.json scripts, etc.)
 
 #### Quality perspective
-- Test infrastructure: framework, runner command, test file locations
+- Test infrastructure: framework, runner command, test file locations and
+  naming patterns — look for `*.test.*`, `*.spec.*`, `test_*.py`, and files
+  under `test/` or `__tests__/`
 - Test coverage: what's tested, what's manual-only
 - Functional capabilities: every feature × how to verify it works
 - Regression requirements: what MUST pass before a release
@@ -177,7 +234,20 @@ List symbols grouped by module/directory:
 
 Use standard chunk frontmatter with rich synonym-expanded `contains:` tags.
 
-If re-onboarding (chunks already exist), follow the archive-then-replace pattern:
+#### Monorepo guidance
+
+If the project has multiple top-level packages or apps (e.g., a `packages/`,
+`apps/`, or `services/` directory with independent `package.json` / `pyproject.toml`
+files), treat each package as a separate chunk group:
+- Write chunks to `{brain_path}/chunks/extracted/project-{safe_project_name}-{package_name}/`
+- Also write a shared overview chunk at `project-{safe_project_name}/chunk-000-overview.md`
+  that summarises the monorepo layout, inter-package relationships, and shared tooling.
+
+#### Re-onboarding
+
+Re-onboarding is triggered **manually** by the user saying "re-onboard this
+project" (or equivalent). It does not run automatically. When re-onboarding
+(chunks already exist), follow the archive-then-replace pattern:
 1. Remove old chunks from index via server API
 2. Archive old chunk directory with `.archived-{timestamp}` suffix
 3. Write new chunks
@@ -228,7 +298,7 @@ returns at depth 1. Put detail after the first paragraph.
 | Article | Depth 0 answers | Depth 1 answers | Depth 2 answers |
 |---------|-----------------|-----------------|-----------------|
 | product.md | "How many features?" | "What are the features?" | "How do I use each one? What are the limits?" |
-| engineering.md | "How many modules/deps?" | "What are the components and how do they connect?" | "What symbols does module X export? How do I extend it?" |
+| engineering.md | "How many modules/deps?" | "What are the components and how do they connect?" | "What symbols does module X export? How do I extend it? What naming/code-style/test/build conventions must I match?" |
 | quality.md | "What's the test coverage?" | "What capabilities need testing?" | "How do I verify capability X works? What are the edge cases?" |
 | operations.md | "How many config files?" | "What can go wrong?" | "How do I fix X? Full troubleshooting playbook." |
 | data.md | "What data sources?" | "What's the schema?" | "What are the constraints? How does the lifecycle work?" |
@@ -248,3 +318,11 @@ Report what was onboarded:
 - Chunks created: {N} (6 perspective-based)
 - Wiki articles: {list of 5 articles}
 - CLI config updated: {file}
+
+## Cross-Platform Notes
+
+- `curl` is cross-platform (Windows 10+) — OK for server API calls.
+- The `.archived-{timestamp}` rename on re-onboarding uses shell on
+  macOS/Linux; on Windows use the equivalent move. Preserve the
+  `.archived-{timestamp}` suffix convention either way.
+- All paths must use forward slashes.

package/skills/wicked-brain-review/SKILL.md

@@ -2,7 +2,7 @@
 name: wicked-brain-review
 description: |
   Browse stored memories with filters on type, tier, and recency. Read-only —
-  use wicked-brain:forget to archive or wicked-brain:agent dispatch consolidate
+  use wicked-brain:forget to archive or wicked-brain:consolidate
   to promote.
 
   Use when: "review my memories", "browse decisions", "what have I stored",
@@ -79,7 +79,7 @@ Age is derived from `indexed_at` relative to now (`Xd` / `Xh`).
 
 After the list, suggest one of:
 - `wicked-brain:forget path=…` to archive a specific entry
-- `wicked-brain:agent dispatch consolidate` to promote patterns and drop expired entries
+- `wicked-brain:consolidate` to promote patterns and drop expired entries
 - `wicked-brain:retag` if many entries have thin `contains:` arrays
 
 ## Notes