agentleFS
Iwan1212Iwan1212/iwan/CLAUDE.md

iwan

CLAUDE.md9 starsChanged 8 months agoMITView on GitHub

Zasada: jedna faza = jeden PR. pnpm turbo test zielone przed mergem.

The file

Files
CLAUDE.md

CLAUDE.md — Kontekst projektu Iwan

Co to jest Iwan

Open-source AI agent Slack. Crawluje wiadomości z kanałów, odpowiada na pytania z kontekstem.

Stack (v0.5 → v1.0 upgrade w toku)

  • Runtime: Node.js 20 LTS
  • Język: TypeScript (strict, NodeNext)
  • Monorepo: pnpm workspaces + Turborepo
  • Slack: @slack/bolt (Socket Mode)
  • AI: @anthropic-ai/sdk (claude-sonnet-4-20250514)
  • Dashboard: React 19 + Vite + Tailwind CSS 4 + TanStack Query
  • API: Express 5 (dashboard REST API)
  • Baza: Supabase (PostgreSQL, full-text search)
  • Hosting: Railway
  • Docker: Node 20 Alpine, multi-stage build (pnpm + turbo)
  • CI: GitHub Actions (pnpm + turbo typecheck/test/build)
  • Testy: Jest 29 (ts-jest), 48 suites, 500+ assertions

Struktura monorepo

apps/bot/         — @iwan/bot (Slack bot + Express API)
apps/dashboard/   — @iwan/dashboard (React SPA admin panel)
packages/shared/  — @iwan/shared (typy + stałe)

Upgrade Plan v0.5 → v1.0 (OpenViktor level)

Cel: 7.5/10+ — kolejność faz: resilience → features → optymalizacja → tooling

FazaCoStatus
0Docker + CI/CDDONE
1TypeScript migrationDONE
2Multi-Provider LLMDONE
3Write Tools (Pipedrive + Slack)DONE
4Redis CacheDONE
5Proactive 2.0 (cron, digest, anomaly)DONE
6Dashboard + MonorepoDONE
7Multi-Workspacedo zrobienia

Zasada: jedna faza = jeden PR. pnpm turbo test zielone przed mergem.

CZEGO JESZCZE NIE UŻYWAMY

  • Nango (integracje zewnętrzne) — planowane v1.0
  • E2B (sandbox) — planowane v2.0
  • Voyage AI (embeddingi) — planowane v1.0
  • pgvector (vector search) — planowane v1.0

Dashboard API

  • Guard: ENABLE_DASHBOARD_API=true (domyślnie off)
  • Auth: Bearer token (DASHBOARD_API_TOKEN)
  • Port: DASHBOARD_API_PORT (default 3100)
  • Endpointy: /api/health, /api/scheduler/jobs, /api/errors, /api/cache/stats, /api/channels, /api/deals/digests, /api/workforce/alerts, /api/config

Zainstalowane zasoby ECC (.claude/)

Skills: docker-patterns, cost-aware-llm-pipeline, backend-patterns, database-migrations, verification-loop, security-review, tdd-workflow, frontend-patterns, api-design, postgres-patterns, deployment-patterns Agents: architect, code-reviewer, tdd-guide, planner, refactor-cleaner, security-reviewer, doc-updater, loop-operator Commands: /plan, /tdd, /code-review, /quality-gate, /build-fix, /verify, /model-route Rules: common/* (9 plików), typescript/* (5 plików)

Integracja Pipedrive CRM

  • Klient API: apps/bot/src/services/pipedrive.ts (search, get, notes, activities)
  • Deal resolver: apps/bot/src/services/dealResolver.ts (Slack kanał → Pipedrive deal, cache w Supabase)
  • Daily digest: apps/bot/src/services/dealDigest.ts (Pn-Pt, automatyczne podsumowania → Pipedrive notes)
  • Narzędzia Claude: search_pipedrive, deal_status, create_deal_note, create_deal_activity, send_slack_message (w tools.ts + toolExecutor.ts)
  • Slash commands: /iwan deal <name>, /iwan deals
  • Backfill: apps/bot/scripts/backfillDeals.js (--days N, --dry-run, --deal "Acme")
  • Knowledge system: apps/bot/knowledge/*.md → injected do LLM prompts (src/services/knowledge.ts)
  • LLM fallback: apps/bot/src/services/openrouter.ts (Anthropic → OpenRouter)
  • Supabase tabele: deal_channel_mappings, deal_digest_state (scripts/seed-deal-tables.sql)

Zasady kodowania

  • Jedna funkcja = jedno zadanie
  • Max 30 linii na funkcję
  • Komentarz PO POLSKU nad każdą funkcją
  • NIE NADPISUJ istniejącego kodu
  • Dodawaj nowe pliki zamiast modyfikować stare
  • Proste nazwy zmiennych (angielskie)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.