feat: Synapse Universal Context Broker (#1)

* Add Synapse design doc and update CLAUDE.md for feature handoff

- docs/synapse-design.md: Full architecture spec for Universal Context Broker
- CLAUDE.md: Synapse section with new key files, resource classes, and conventions
- Branch: feature/synapse

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add Synapse phases 1-3: schema, resource classes, and tests

- Add SynapseEntry table to schema.graphql with HNSW vector index
- Add SynapseEntry, SynapseSearch, SynapseIngest, SynapseEmit to resources.js
- Add classifySynapseEntry, synapseparsers, synapseEmitters helpers
- Update existing tests to include SynapseEntry in harperdb mock
- Add 4 new Synapse test files (77 tests total, all pass)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add Synapse CLI (bin/synapse.js) with sync, emit, search, watch, status

- Zero-dependency Node.js CLI using process.argv and fetch
- sync: discovers CLAUDE.md, .cursor/rules/*.mdc, .windsurf/ and POSTs to SynapseIngest
- emit: queries SynapseEmit and prints or writes files (--write flag)
- search: semantic search via SynapseSearch with --limit and --type filters
- watch: fs.watch with 2s debounce for auto-sync on file changes
- status: shows entry counts by type and source from a search sample
- Add "bin" field to package.json
- Add SYNAPSE_ENDPOINT, SYNAPSE_PROJECT, SYNAPSE_AUTH to .env.example

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add Synapse documentation (Phase 5)

- README.md: add Synapse section with architecture diagram, context types,
  CLI examples, API endpoints table, project structure, and env vars
- docs/architecture.md: add Synapse data flow diagram and SynapseEntry schema table
- CLAUDE.md: update key files list, test count (35 → 77), remove branch qualifier

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add linting, formatting, and release tooling

- dprint: code formatter (tabs, single quotes, useBraces:always) for JS, JSON, Markdown, GraphQL
- oxlint: fast linter with test-file overrides for no-new-array and require-yield
- semantic-release: automated versioning and GitHub releases from conventional commits
- commitlint: enforce conventional commit format
- GitHub Actions: ci.yml (lint+test on all branches) and release.yml (release on main)
- New scripts: format:check, format:fix, format:staged, lint:check, lint:fix, test:watch, test:coverage, commitlint
- Fix lint warnings in resources.js: prefix unused embedding and projectId params with _

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: address Synapse review issues

- Add VALID_EMIT_TARGETS set; reject slack/manual as emit targets, accept markdown
- Push single-type filter to server-side search conditions in SynapseEmit
- Deduplicate ingested entries using deterministic content hash as record ID
- Fix parseClaudeCode/parseWindsurf to preserve preamble before first ## heading
- Update engines.node to >=22.0.0 (required for --experimental-test-module-mocks)
- Combine table destructures into single line
- Add beforeEach mock resets to synapse-search and synapse-emit test files
- Broaden CLI status command query; document as approximate

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: daniel-harperdb

.env.example

@@ -24,3 +24,9 @@ VOYAGE_API_KEY=pa-xxxxx
 # See docs/slack-app-setup.md for step-by-step instructions
 SLACK_SIGNING_SECRET=xxxxx
 SLACK_BOT_TOKEN=xoxb-xxxxx
+
+# --- Synapse CLI (for context broker commands) ---
+# See docs/synapse-design.md for usage instructions
+SYNAPSE_ENDPOINT=https://your-cluster.your-org.harperfabric.com
+SYNAPSE_PROJECT=your-project-name
+SYNAPSE_AUTH=Basic dXNlcjpwYXNz

.github/workflows/ci.yml

@@ -0,0 +1,30 @@
+name: CI
+
+on:
+  push:
+    branches: ['**']
+  pull_request:
+    branches: [main]
+
+jobs:
+  test:
+    name: Lint & Test
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version-file: .nvmrc
+          cache: npm
+
+      - run: npm ci
+
+      - name: Format check
+        run: npm run format:check
+
+      - name: Lint
+        run: npm run lint:check
+
+      - name: Test
+        run: npm test

.github/workflows/release.yml

@@ -0,0 +1,35 @@
+name: Release
+
+on:
+  push:
+    branches: [main]
+
+permissions:
+  contents: write
+  issues: write
+  pull-requests: write
+
+jobs:
+  release:
+    name: Release
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          persist-credentials: false
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version-file: .nvmrc
+          cache: npm
+
+      - run: npm ci
+
+      - name: Test
+        run: npm test
+
+      - name: Release
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: npx semantic-release

.oxlintrc.json

@@ -0,0 +1,16 @@
+{
+	"$schema": "./node_modules/oxlint/configuration_schema.json",
+	"ignorePatterns": [
+		"node_modules/"
+	],
+	"rules": {},
+	"overrides": [
+		{
+			"files": ["test/**"],
+			"rules": {
+				"unicorn/no-new-array": "off",
+				"require-yield": "off"
+			}
+		}
+	]
+}

.releaserc.json

@@ -0,0 +1,33 @@
+{
+	"branches": [
+		"main"
+	],
+	"plugins": [
+		[
+			"@semantic-release/commit-analyzer",
+			{
+				"preset": "conventionalcommits",
+				"releaseRules": [
+					{ "breaking": true, "release": "major" },
+					{ "type": "feat", "release": "minor" },
+					{ "type": "*", "release": "patch" }
+				],
+				"parserOpts": {
+					"noteKeywords": ["BREAKING CHANGE", "BREAKING CHANGES"]
+				}
+			}
+		],
+		"@semantic-release/release-notes-generator",
+		[
+			"@semantic-release/npm",
+			{ "npmPublish": false }
+		],
+		[
+			"@semantic-release/git",
+			{
+				"assets": ["package.json", "package-lock.json"]
+			}
+		],
+		"@semantic-release/github"
+	]
+}

CLAUDE.md

@@ -14,23 +14,47 @@ An agent-agnostic AI memory system using Harper Fabric as the vector database an
 
 ## Key Files
 
-- `resources.js` - Core application: SlackWebhook, MemorySearch, MemoryTable classes + classifyMessage, generateEmbedding, verifySlackSignature helpers
-- `schema.graphql` - Memory table with HNSW vector index (no @export since we extend it)
+- `resources.js` - Core application: SlackWebhook, MemorySearch, MemoryTable + all Synapse resource classes + helpers
+- `schema.graphql` - Memory and SynapseEntry tables with HNSW vector indexes (no @export since we extend them)
 - `config.yaml` - Harper app config (loadEnv, REST, schema, resource files)
 - `.env.example` - All required environment variables documented
+- `bin/synapse.js` - Synapse CLI: sync, emit, search, watch, status commands
 
 ## Development
 
 ```bash
 npm run dev    # Start locally on port 9926
-npm test       # Run all 35 tests
+npm test       # Run all 77 tests
 npm run deploy # Deploy to Harper Fabric
 ```
 
 ## Architecture
 
 Slack webhook -> classify (Claude) + embed (Voyage AI) -> store in Memory table -> query via MCP from Claude Desktop/Cursor.
 
+## Synapse
+
+Universal Context Broker — ingests development context from any AI tool (Claude Code, Cursor, Windsurf, Copilot) and emits it in any other tool's native format. Full design spec: `docs/synapse-design.md`.
+
+### New Key Files
+
+- `bin/synapse.js` - CLI: sync, emit, search, watch, status commands
+- `test/synapse-*.test.js` - Tests for classify, search, ingest, emit
+
+### New Resource Classes (in resources.js)
+
+- `SynapseEntry` - Table extension (strips embeddings, same pattern as MemoryTable)
+- `SynapseSearch` - Semantic search with mandatory `projectId` scoping
+- `SynapseIngest` - Parses tool-native formats into SynapseEntry records
+- `SynapseEmit` - Formats SynapseEntry records into tool-native output
+
+### Conventions
+
+- SynapseEntry table follows same patterns as Memory (HNSW vector index, classification via Claude Haiku, embeddings via Voyage AI)
+- Use renamed import: `const { SynapseEntry: SynapseEntryBase } = tables;`
+- All Synapse queries must filter on `projectId`
+- Default status filter is `active` (excludes superseded/archived)
+
 ## Agent Skills
 
 Skills from `harperfast/skills` are tracked in `skills-lock.json` and installed into `.agents/skills/` (git-ignored). Refer to the relevant skill rules when making changes:

CONTRIBUTING.md

@@ -57,12 +57,12 @@ npm test
 
 Tests live in `test/` alongside the code they cover:
 
-| File | Covers |
-|------|--------|
-| `classify.test.js` | `classifyMessage()` |
-| `embedding.test.js` | `generateEmbedding()` |
-| `webhook.test.js` | `SlackWebhook`, `verifySlackSignature()` |
-| `search.test.js` | `MemorySearch` |
+| File                | Covers                                   |
+| ------------------- | ---------------------------------------- |
+| `classify.test.js`  | `classifyMessage()`                      |
+| `embedding.test.js` | `generateEmbedding()`                    |
+| `webhook.test.js`   | `SlackWebhook`, `verifySlackSignature()` |
+| `search.test.js`    | `MemorySearch`                           |
 
 ## Submitting a Pull Request
 
@@ -76,6 +76,7 @@ Please keep PRs focused — one feature or fix per PR.
 ## Reporting Issues
 
 Open an issue on GitHub with:
+
 - A clear description of the problem
 - Steps to reproduce
 - Your Node.js version (`node --version`)