@stackmemoryai/stackmemory 1.0.1 → 1.2.1

package/scripts/gepa/generations/gen-001/variant-a.md

@@ -0,0 +1,166 @@
+# ProvenantAI
+
+## Reference Docs
+```
+AGENTS.md                                # TDD workflow, checklists, guardrails
+PROMPT_PLAN.md                           # 20 implementation prompts
+docs/STYLE.md                            # Design system (Hatchet + Outliner)
+docs/business/ONE_PAGER.md               # Executive summary
+DEV_SPEC.md                              # Developer spec
+docs/reference/PROJECT.md                # Quick reference
+docs/business/VISION.md                  # Product vision + self-learning + model router
+docs/architecture/SYSTEM_INTEGRATION.md  # System connections (events, flows)
+docs/architecture/HEARTBEAT_DESIGN.md    # Task health + recovery
+docs/nudge-engine-design.md              # Proactive scanner alerts
+docs/architecture/WEBHOOK_SYSTEM_DESIGN.md  # External event publishing
+docs/VALUES.md                           # Company values
+```
+
+## Commands
+```bash
+npm run dev|test|lint|migrate           # Development tasks
+docker-compose up -d                    # Start local DBs
+railway up                              # Deploy to Railway
+```
+
+## Stack
+Node.js + Express + PostgreSQL + Redis | Railway | Stripe + Salesforce + QuickBooks
+
+## Core Services (src/core/)
+monitoring-service | cache-service | queue-service | master-agent | api-validation
+
+## Directory Structure
+```
+src/
+  api/ core/ features/ shared/ integrations/
+docs/ scripts/ docker/
+```
+
+## Git Workflow
+- **NEVER add Co-Authored-By lines** to commits
+- Pre-commit: runs lint + test suite
+- Commit format: `type(scope): message`
+- Deploy: `railway up --detach`
+
+## Architecture
+- **Provenance tracking**: every data point includes source + timestamp + lineage
+- **Multi-tenant**: container isolation per org
+- **Investigation replays**: stored in `data/investigation-replays/`
+- **StackMemory**: session/entity context layer (repurpose when KG lands)
+
+## Key Implementation Files
+```
+src/graph/client.ts                     # MCP client for Graphiti
+src/graph/service.ts                    # Graph ingest + enrich + search
+src/llm/adapter.js                      # LLM streaming + SYSTEM_PROMPT
+src/routes/query.js                     # Query endpoint + KG enrichment
+src/auth/auth.middleware.js             # Clerk + API key + test mode
+src/nudge/rule-engine.js                # 13 default rules + CRUD + eval
+src/nudge/nudge-engine.js               # Lifecycle + state machine + events
+src/nudge/delivery.js                   # Slack Block Kit + delivery
+src/integrations/slack/app.js           # OAuth v2 + token management
+scripts/create-stripe-products.js       # Stripe product creation
+scripts/seed-demo-data.js               # Demo nudge data
+dashboard-app/src/pages/pricing/        # 4-tier pricing UI
+```
+
+## Current State
+- **20 prompts complete** (PROMPT_PLAN.md)
+- **80 test suites, 1547 tests** - all passing (37 unit, 30 core, 13 integrations)
+- **KG integrated** - commit 917f31e (FalkorDB + Graphiti MCP)
+- **Published**: @provenantai/cli@1.0.0
+- **Deployed**: Railway (health OK, DB connected)
+- **Bundle**: 236KB main + 346KB vendor (lazy routes)
+- **Pricing**: Free | Growth $999-1499 | Scale $2999-4499 | Enterprise custom
+- **Slack**: wired in index.js (conditional on SLACK_CLIENT_ID)
+- **Dashboard**: SlackSettings + Nudges page + E2E smoke tests
+
+## Knowledge Graph
+- **FalkorDB**: `falkordb/falkordb:latest` port 6380 (Redis-compatible)
+- **Graphiti**: `zepai/knowledge-graph-mcp:standalone` port 8100
+- **Protocol**: MCP Streamable HTTP (JSON-RPC + SSE) - NOT REST
+- **Ingestion**: fire-and-forget via `add_memory`, extraction runs in background
+- **Config**: `config/graphiti-falkordb.yaml` - Claude Haiku for LLM
+- **Performance**: 34ms avg, $0.00056/event
+- **Query integration**: GraphService.enrichRetrieval() when GRAPHITI_URL set
+- **Graph context**: prepended as first context_document, graceful fallback
+
+## Stripe Products (LIVE)
+```
+Growth:  prod_TyNYCJmlKbMdlz
+  Monthly $1,499:    price_1T0Qn8BjaxUVbh5VjC7BP0cS
+  Quarterly $2,997:  price_1T0Qn9BjaxUVbh5VYmugdPrF
+Scale:   prod_TyNY5SGKvVlbpo
+  Monthly $4,499:    price_1T0Qn9BjaxUVbh5ViK6vRmgn
+  Quarterly $8,997:  price_1T0Qn9BjaxUVbh5V1mBcTAJ0
+```
+Price IDs in `.env` + `dashboard-app/.env` (VITE_STRIPE_PRICE_*)
+
+## Clerk Auth
+- **Current**: test keys (pk_test_*, sk_test_*)
+- **Production setup**: `bash scripts/setup-clerk-production.sh`
+- **Webhook**: /api/webhooks/clerk (Svix verification required in prod)
+- **Status**: code production-ready, test mode blocked in production
+
+## Dashboard
+- **Chat**: `/api/v1/query` with SSE streaming
+- **Sidebar**: toggleable, time-grouped, backend persistence
+- **localStorage**: `provenantai:conversationId`, `sidebarOpen`, `onboarded`
+- **Vite**: `base: '/app/'` + BrowserRouter `basename="/app"`
+- **Dev auth**: `X-Test-Mode: true` bypasses Clerk
+- **Express**: `node src/index.js` on PORT=8080 (3000 = Rails)
+- **Build**: `bash -c 'cd dashboard-app && npx vite build'`
+
+## Slack Integration
+- **OAuth flow**: GET /api/integrations/slack/install → consent → callback
+- **DB**: `slack_installations` table (migration 030) - per-org bot tokens
+- **Service**: `src/integrations/slack/app.js` - createSlackAppService
+- **Delivery**: `src/nudge/delivery.js` - nudge:created listener, Block Kit formatter
+- **Routes**: install, callback, status, channel config, revoke
+- **Env**: SLACK_CLIENT_ID, SLACK_CLIENT_SECRET, SLACK_REDIRECT_URI
+- **Demo data**: `node scripts/seed-demo-data.js` (10 nudges)
+
+## Architecture Patterns
+- **DI**: all route factories accept `deps` objects
+- **Mocks**: MockPineconeAdapter, MockGmailConnectorClient, stub embeddings
+- **Queues**: Bull (not BullMQ)
+- **KMS**: AES-256-GCM shim via LOCAL_KMS_KEY
+- **Streaming**: adapter.streamLLM() → query/stream → ReadableStream
+- **Graph**: optional graphService dep, null when GRAPHITI_URL unset
+
+## Testing
+- **Framework**: Jest + SWC (`@swc/jest`), `@jest/globals` imports
+- **Mocks**: `mockDb = { query: jest.fn() }` via DI
+- **Fetch**: `global.fetch = mockFetch` + `jsonResponse()` helper
+- **Pre-commit**: lint + parallel test suite via concurrently
+- **Timeout guard**: Promise.race + .unref() timer in test/setup.js
+- **Supertest**: pass Express app directly (NOT server) - avoids port contention
+- **Redis**: skip eager connect in NODE_ENV=test; disconnect in afterAll
+
+### Parallel Test Execution
+```bash
+npm test                                 # 3 parallel processes (~9s vs 18s)
+test:unit       (37 suites, 615 tests)  # test/ dir
+test:core       (30 suites, 643 tests)  # src/(api|auth|billing|core|...)
+test:integrations (13 suites, 289 tests)  # src/integrations/
+test:all        (single-process, 4 workers)  # fallback
+```
+**After code changes**: run targeted sub-suite in background via Bash `run_in_background`
+
+## GTM Content
+```
+docs/content/linkedin-pillar1-drafts.md          # 5 problem posts (ready)
+docs/content/gtm-launch-materials.md             # Schedule + Loom + outreach
+docs/business/CONTENT_INBOUND_STRATEGY.md        # Content/inbound playbook
+docs/business/FRACTIONAL_CMO_AGENT.md            # CMO Agent product brief
+```
+
+## Next Actions
+1. Commit: SlackSettings, Nudges polish, smoke tests
+2. Clerk: swap to production keys (requires Clerk dashboard action)
+3. Slack: create app at api.slack.com, run migration 030
+4. Content: schedule posts in Typefully, record Loom demo
+5. Design partners: outreach to 3-5 ICP companies
+6. Attribution: add Google Ads + Meta connectors
+
+<!-- Sync docs/reference/PROJECT.md when verbose docs change -->

package/scripts/gepa/generations/gen-001/variant-b.md

@@ -0,0 +1,237 @@
+# CLAUDE.md [compact]
+
+## Refs
+```
+~/.claude/MCP.md|PERSONAS.md|STACKMEMORY.md
+~/.claude/agent_docs/*.compact.md (on-demand)
+```
+
+## Commands
+```
+build|lint|lint:fix|test|test:run
+git status|diff|log --oneline -10
+npx tsc --noEmit|npm run format
+```
+
+## Core
+[PRINCIPLE]code>docs|simple→complex|security first|evidence-based
+[COMM]concise|symbols>prose|bullets>paragraphs|<4 lines
+[WORKFLOW]TodoWrite(3+)→execute→update
+[GIT]clean commits|type(scope): message|feature/|fix/|chore/ prefix
+[STACK]React/TS/Vite|Node/Express/PG|Git/ESLint/Jest
+
+## Think
+[NONE]single file,<10 lines
+[THINK]multi-file,standard|~4K
+[HARD]architecture,complex|~10K
+[ULTRA]critical redesign|~32K
+
+## Critical [C:10]
+[SECURITY]
+  NEVER:commit secrets|exec untrusted|expose PII|force push
+  ALWAYS:validate input|parameterized queries|hash passwords
+  BLOCK:~/.ssh|~/.aws|/api[_-]?key|token|secret/i
+
+<example>
+❌ BAD:
+```js
+db.query(`SELECT * FROM users WHERE id = ${userId}`)
+```
+✅ GOOD:
+```js
+db.query('SELECT * FROM users WHERE id = $1', [userId])
+```
+</example>
+
+[ESM]add .js to relative imports|use ts-node-lint-fixer agent on ERR_MODULE_NOT_FOUND
+
+<example>
+❌ BAD:
+```ts
+import { foo } from './bar'
+```
+✅ GOOD:
+```ts
+import { foo } from './bar.js'
+```
+</example>
+
+[ERROR]return undefined>throw|log+continue>crash|filter nulls
+
+<example>
+❌ BAD:
+```js
+function getUser(id) {
+  if (!id) throw new Error('No ID')
+  return users.find(u => u.id === id)
+}
+```
+✅ GOOD:
+```js
+function getUser(id) {
+  if (!id) return undefined
+  return users.find(u => u.id === id)
+}
+```
+</example>
+
+[CODE]no emojis|comments only complex logic|short names
+
+<example>
+❌ BAD:
+```js
+// This function gets the user from the database
+function getUserFromDatabase(userId) { ... }
+```
+✅ GOOD:
+```js
+function getUser(id) { ... }
+// Only comment complex parts:
+const hash = await bcrypt.hash(pwd, 10) // 10 rounds for security/perf balance
+```
+</example>
+
+## High [H:8-9]
+[EFFICIENCY]do>explain|action>ceremony|parallel>sequential
+
+<example>
+❌ BAD: "I'll read file A, then file B, then file C..."
+✅ GOOD: [Read A, Read B, Read C in parallel]
+</example>
+
+[GIT]status→branch→fetch→pull --rebase|type(scope): message
+
+<example>
+✅ GOOD commit messages:
+- feat(auth): add JWT token refresh
+- fix(api): handle null responses
+- chore(deps): bump express to 4.18
+</example>
+
+[RECOVERY]try alt→explain→suggest next|never silent fail
+
+<example>
+❌ BAD: Test fails → retry same command → retry again
+✅ GOOD: Test fails → check logs → try different approach → explain issue
+</example>
+
+[SESSION]track edits/corrections/paths|cache versions/locations
+
+## Standards
+[TASK]TodoWrite 3+|one in_progress|update immediate
+
+<example>
+✅ GOOD workflow:
+1. User asks to add feature
+2. TodoWrite: ["Design API", "Implement handler", "Add tests", "Update docs"]
+3. TaskUpdate task1 → in_progress
+4. Complete work
+5. TaskUpdate task1 → completed
+6. TaskUpdate task2 → in_progress
+</example>
+
+[DESIGN]KISS|YAGNI|SOLID|<20 lines/fn|<5 complexity
+
+<example>
+❌ BAD:
+```js
+function processUserDataWithValidationAndTransformation(user, options) {
+  // 50 lines of mixed concerns
+}
+```
+✅ GOOD:
+```js
+function validateUser(user) { ... }     // 8 lines
+function transformUser(user) { ... }   // 6 lines
+function processUser(user) {           // 3 lines
+  const valid = validateUser(user)
+  return valid ? transformUser(user) : null
+}
+```
+</example>
+
+[FILES]read before write|edit>write|no docs unless asked
+
+<example>
+❌ BAD: User asks to update function → Write entire new file
+✅ GOOD: User asks to update function → Read file → Edit specific function
+</example>
+
+[VALIDATE]lint→test→build→run|never assume success
+
+<example>
+✅ GOOD workflow:
+1. Edit code
+2. `npm run lint` → check output
+3. `npm test` → verify passing
+4. `npm run build` → ensure clean build
+5. Only then mark task complete
+</example>
+
+[COVERAGE]maintain or improve test coverage|no untested code paths
+
+<example>
+❌ BAD: Add new route without tests
+✅ GOOD: Add new route + unit test + integration test
+</example>
+
+## Style
+[OUTPUT]concise|structured|actionable
+
+<example>
+❌ BAD: "I've made some changes to the authentication system to improve security..."
+✅ GOOD: "Added JWT refresh tokens in src/auth/tokens.js:45"
+</example>
+
+[PUSHBACK]"Simpler: X"|"Risk: Y"|"Consider: Z"
+
+<example>
+User: "Add Redis caching to all endpoints"
+✅ GOOD response: "Risk: premature optimization. Consider: profile first, cache hot paths only"
+</example>
+
+[QUESTIONS]1-3 clarifying|one at a time|no time estimates
+
+<example>
+❌ BAD: "This will take 2-3 hours. Should I add error handling, logging, tests, docs, and type safety?"
+✅ GOOD: "Add error handling for network failures?"
+</example>
+
+## Summary Format
+```
+Session: actual vs estimated|variance %
+Completed: N/M tasks|files modified|commits
+Outcomes: deliverables|blockers|next actions
+```
+
+<example>
+✅ GOOD summary:
+```
+Completed: 3/4 tasks | 7 files | 2 commits
+Outcomes: Auth refresh implemented, tests passing
+Blockers: Redis connection requires env var
+Next: Add REDIS_URL to .env, deploy to staging
+```
+</example>
+
+## Auto-Activate
+[FILES]*.tsx→frontend|*.sql→data|Docker→devops|*.test→qa
+[KEYWORDS]bug/error→debugger|optimize→perf|secure→security
+
+<example>
+User shares error.tsx → automatically apply frontend patterns
+User mentions "slow query" → automatically consider performance context
+</example>
+
+## Expand (read on match)
+[AGENTIC]multi-agent→AGENTIC_CODING.compact.md
+[CONTEXT]token budget→CONTEXT_MANAGEMENT.compact.md
+[TOOLS]parallel tools→TOOL_USE.compact.md
+[HORIZON]multi-session→LONG_HORIZON.compact.md
+[PROMPTS]prompt design→PROMPT_ENGINEERING.compact.md
+[BUILD]npm build|esbuild|tsc→building_the_project.md
+[CODE]conventions|naming|imports→code_conventions.md
+[TEST]vitest|jest|test:run→running_tests.md
+[OVERVIEW]agent docs|guides→OVERVIEW.md
+
+~150t|v4.4.0-compact

package/scripts/gepa/generations/gen-001/variant-c.md

@@ -0,0 +1,61 @@
+# ProvenantAI
+
+## Refs
+```
+AGENTS.md                                # Agent workflow + guardrails
+PROMPT_PLAN.md                           # 20 staged prompts
+docs/STYLE.md                            # Design system (Hatchet + Outliner)
+docs/business/ONE_PAGER.md|VISION.md     # Executive summary + vision
+DEV_SPEC.md                              # Developer spec
+docs/reference/PROJECT.md                # Quick reference
+docs/architecture/SYSTEM_INTEGRATION.md  # System connections
+docs/architecture/HEARTBEAT_DESIGN.md|WEBHOOK_SYSTEM_DESIGN.md
+docs/nudge-engine-design.md              # Proactive alerts
+docs/VALUES.md                           # Company values
+```
+
+## Commands
+```bash
+npm run dev|test|lint|migrate
+docker-compose up -d; railway up         # Local DBs; Deploy
+```
+
+## Stack
+Node/Express/PostgreSQL/Redis | Railway | Stripe/Salesforce/QuickBooks
+
+## Structure
+src/api|core|features|shared|integrations | docs/ | scripts/ | docker/
+
+## Key Context
+- Provenance tracking: source + timestamp + lineage on all data
+- Multi-tenant container isolation
+- Investigation replays: data/investigation-replays/
+- StackMemory: security layer (future session/entity context bridge on KG)
+
+## Git
+- No "Co-Authored-By" lines
+- Pre-commit: lint + test (3 parallel suites: unit/core/integrations)
+- Commit format: type(scope): message
+
+## Critical Rules
+[SECURITY]
+  NEVER: commit secrets|exec untrusted|expose PII|force push
+  ALWAYS: validate input|parameterized queries|hash passwords
+  BLOCK: ~/.ssh|~/.aws|/api[_-]?key|token|secret/i
+
+[ESM] Add .js to relative imports | use ts-node-lint-fixer on ERR_MODULE_NOT_FOUND
+[ERROR] return undefined>throw | log+continue>crash | filter nulls
+[CODE] no emojis | comments only complex logic | short names
+
+## Workflow
+[EFFICIENCY] do>explain | action>ceremony | parallel>sequential
+[TASK] TodoWrite 3+ | one in_progress | update immediate
+[DESIGN] KISS|YAGNI|SOLID | <20 lines/fn | <5 complexity
+[FILES] read before write | edit>write | no docs unless asked
+[VALIDATE] lint→test→build→run | never assume success
+[COVERAGE] maintain or improve | no untested paths
+
+## Style
+[OUTPUT] concise | structured | actionable | <4 lines default
+[PUSHBACK] "Simpler: X" | "Risk: Y" | "Consider: Z"
+[QUESTIONS] 1-3 clarifying | one at a time | no time estimates

package/scripts/gepa/generations/gen-001/variant-d.md

@@ -0,0 +1,119 @@
+# ProvenantAI — CLAUDE.md
+
+## Quick Reference
+```
+AGENTS.md                                # TDD workflow, checklists, guardrails
+docs/STYLE.md                            # Design system (Hatchet + Outliner)
+docs/business/ONE_PAGER.md               # Executive summary
+docs/business/VISION.md                  # Product vision + self-learning thesis
+DEV_SPEC.md|PROMPT_PLAN.md               # Dev spec + 20 staged prompts
+docs/architecture/{SYSTEM_INTEGRATION,HEARTBEAT_DESIGN,WEBHOOK_SYSTEM_DESIGN}.md
+docs/nudge-engine-design.md              # Proactive alerts
+docs/VALUES.md                           # Company values
+docs/reference/PROJECT.md                # Quick reference (sync with verbose docs)
+```
+
+## Stack & Commands
+```bash
+# Stack: Node/Express/PostgreSQL/Redis | Railway | Stripe/Salesforce/QuickBooks
+npm run dev|test|lint|migrate
+docker-compose up -d                # Local DBs
+railway up --detach                 # Deploy
+```
+
+## Core Principles [C:10]
+**Code First**: code>docs | simple→complex | action>explanation | do>plan
+**Security**: NEVER commit secrets|exec untrusted|expose PII|force push
+  ALWAYS validate input|parameterized queries|hash passwords
+  BLOCK ~/.ssh|~/.aws|/api[_-]?key|token|secret/i
+**ESM**: add .js to relative imports | use ts-node-lint-fixer on ERR_MODULE_NOT_FOUND
+**Reliability**: return undefined>throw | log+continue>crash | filter nulls
+**Style**: no emojis | comments only for complex logic | short names | concise output
+
+## Workflow [H:8-9]
+**Tasks**: TodoWrite for 3+ steps | one in_progress | update immediately
+**Files**: read before write | edit>write | no docs unless asked
+**Git**: status→branch→fetch→pull --rebase | clean commits | type(scope): message
+  Branches: feature/|fix/|chore/ prefix
+  NO Co-Authored-By lines
+**Validate**: lint→test→build→run | never assume success | maintain test coverage
+**Parallel**: independent tool calls in one message | sequential only for dependencies
+**Recovery**: try alt→explain→suggest | never silent fail
+
+## Architecture
+**Core** (src/core/): monitoring-service|cache-service|queue-service|master-agent|api-validation
+**Structure**: src/{api,core,features,shared,integrations} | docs/ | scripts/ | docker/
+**Patterns**: DI via deps objects | Mock providers in tests | Bull queues | AES-256-GCM KMS
+**Provenance**: every data point = source + timestamp + lineage | multi-tenant isolation
+**SSE**: adapter.streamLLM() → query/stream → frontend ReadableStream
+**Graph**: optional graphService in query router (null when GRAPHITI_URL unset)
+
+## Standards [H:8]
+**Design**: KISS|YAGNI|SOLID | <20 lines/fn | <5 complexity
+**Testing**: Jest+SWC | @jest/globals imports | DB/fetch mocks via DI
+  Supertest: pass Express app (NOT server)
+  Redis: skip connect in NODE_ENV=test
+  afterAll: Promise.race + .unref() timer
+  Parallel suites: test:unit|test:core|test:integrations (~9s vs 18s)
+  **After code changes**: run targeted sub-suite in background
+**Coverage**: maintain or improve | no untested paths
+
+## Key Context
+- All 20 PROMPT_PLAN.md prompts complete | 80 suites, 1547 tests passing
+- KG: FalkorDB port 6380 + Graphiti MCP port 8100 | MCP protocol (not REST)
+- @provenantai/cli@1.0.0 published | Railway deployed | Bundle: 236KB+346KB
+- Stripe: Growth (prod_TyNYCJmlKbMdlz) $1499/mo | Scale (prod_TyNY5SGKvVlbpo) $4499/mo
+- Clerk: test keys active | production validator: scripts/setup-clerk-production.sh
+- Dashboard: /app/ base | SSE to /api/v1/query | Sidebar in localStorage
+- Slack: OAuth v2 (migration 030) | delivery.js listens nudge:created | demo: seed-demo-data.js
+
+## Key Files
+```
+src/graph/{client.ts,service.ts}         # MCP client + graph logic
+src/llm/adapter.js                       # SYSTEM_PROMPT + streaming
+src/routes/query.js                      # Query + KG enrichment
+src/auth/auth.middleware.js              # Clerk + API key + test mode
+src/nudge/{rule-engine,nudge-engine,delivery}.js  # Rules + lifecycle + Slack
+src/integrations/slack/app.js            # OAuth v2 + token mgmt
+scripts/{create-stripe-products,seed-demo-data}.js
+dashboard-app/src/pages/{pricing/PricingPage,checkout/*}.tsx
+docs/content/{linkedin-pillar1-drafts,gtm-launch-materials}.md
+docs/business/{CONTENT_INBOUND_STRATEGY,FRACTIONAL_CMO_AGENT}.md
+```
+
+## Design System (docs/STYLE.md)
+- Layout: 208px sidebar | inset shadow main | shadow-soft variants
+- Type: 10-24px (no text-lg+ in app) | tabular-nums | Inter/SF Mono | font-mono headers
+- Color: slate/zinc | brand-600 blue | rose accent CTAs
+- Anti-patterns: no @apply sprawl | no inline hex | no shadow-md+ | max rounded-xl
+
+## Output Format
+**Session Summary**:
+```
+Actual vs estimated | variance %
+Completed: N/M tasks | files modified | commits
+Deliverables | blockers | next actions
+```
+**Communication**: concise | bullets>paragraphs | <4 lines | structured
+**Pushback**: "Simpler: X" | "Risk: Y" | "Consider: Z"
+**Questions**: 1-3 max | one at a time | no time estimates
+
+## Token Budget (Think Mode)
+[NONE] single file, <10 lines
+[THINK] multi-file, standard | ~4K
+[HARD] architecture, complex | ~10K
+[ULTRA] critical redesign | ~32K
+
+## Auto-Activate
+[FILES] *.tsx→frontend | *.sql→data | Docker→devops | *.test→qa
+[KEYWORDS] bug/error→debugger | optimize→perf | secure→security
+
+## Next Priorities (Memory)
+- Commit uncommitted agent work (SlackSettings, Nudges polish, smoke tests)
+- Clerk: swap to production keys (user action in dashboard)
+- Slack: create app at api.slack.com, run migration 030
+- Content: schedule pillar posts in Typefully, record Loom demo
+- Design partners: outreach to 3-5 ICP companies
+- Ad platform connectors: Google Ads / Meta for attribution story
+
+<!-- Update docs/reference/PROJECT.md when verbose docs change -->

package/scripts/gepa/results/eval-1-baseline.json

@@ -0,0 +1,41 @@
+{
+  "variant": "baseline",
+  "generation": 1,
+  "results": [
+    {
+      "taskId": "eval-001",
+      "taskName": "simple_function",
+      "passed": false,
+      "duration": 6,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-002",
+      "taskName": "refactor_code",
+      "passed": false,
+      "duration": 5,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-003",
+      "taskName": "fix_bug",
+      "passed": false,
+      "duration": 5,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-004",
+      "taskName": "add_feature",
+      "passed": false,
+      "duration": 5,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-005",
+      "taskName": "code_review",
+      "passed": false,
+      "duration": 4,
+      "output": "/bin/sh: timeout: command not found\n"
+    }
+  ]
+}

package/scripts/gepa/results/eval-1-variant-a.json

@@ -0,0 +1,41 @@
+{
+  "variant": "variant-a",
+  "generation": 1,
+  "results": [
+    {
+      "taskId": "eval-001",
+      "taskName": "simple_function",
+      "passed": false,
+      "duration": 5,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-002",
+      "taskName": "refactor_code",
+      "passed": false,
+      "duration": 10,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-003",
+      "taskName": "fix_bug",
+      "passed": false,
+      "duration": 7,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-004",
+      "taskName": "add_feature",
+      "passed": false,
+      "duration": 6,
+      "output": "/bin/sh: timeout: command not found\n"
+    },
+    {
+      "taskId": "eval-005",
+      "taskName": "code_review",
+      "passed": false,
+      "duration": 5,
+      "output": "/bin/sh: timeout: command not found\n"
+    }
+  ]
+}