Iwan

Open Source AI Agent for Slack — with Admin Dashboard

Mention @Iwan in Slack — it searches your Slack history, Notion, Pipedrive CRM, Workforce Planner, Calamari and Google Calendar, then answers using Claude with full context. Ships with a React admin dashboard for ops.

Getting Started · How It Works · Integrations · Dashboard · Slash Commands · Roadmap

What Iwan Does

Iwan is a Slack bot + admin dashboard. The bot connects to six data sources — Slack message history, Notion, Pipedrive CRM, Workforce Planner, Calamari (PTO) and Google Calendar — and uses Claude (Sonnet 4.5 + Haiku 4.5) to answer questions with context from all of them.

You ask a question in Slack. In tool-use mode Claude decides which sources to query and calls them as tools; in legacy mode all sources are queried in parallel. The whole thing takes 2–5 seconds.

Iwan also runs scheduled jobs (cron) — daily Pipedrive deal digests, weekly team allocation summaries, workforce anomaly detection, channel digests — and exposes everything through a REST API consumed by a React admin dashboard.

Example interaction:

You:   @Iwan kto wolny w marcu?

Iwan:  Na podstawie Workforce Planner — wolni w marcu:

       Backend:
       - Alex Johnson — 0% utilization
       - Sam Chen — 12% utilization

       PM:
       - Maria Torres — 0% utilization

       Razem: 4 osoby (z utilizacją <30%)

Current State (v0.5)

Phase	What	Status
0	Docker + CI/CD	✅
1	TypeScript migration	✅
2	Multi-Provider LLM (OpenRouter fallback)	✅
3	Write Tools (Pipedrive + Slack)	✅
4	Redis Cache	✅
5	Proactive 2.0 (cron, digest, anomaly)	✅
6	Dashboard + Monorepo	✅
7	Multi-Workspace	🔜

Known limitations:

Slash commands are Polish-only — /iwan szukaj, /iwan kto-wolny etc. English aliases planned for v0.6.
Search is full-text, not semantic — Slack history and Notion use keyword matching via Supabase. Voyage AI + pgvector planned for v0.7.
Single workspace only — no multi-tenant support yet (phase 7).
JWT tokens are in-memory — Workforce Planner re-authenticates on restart.

How It Works

@Iwan "kto wolny w marcu?"
    │
    ├── Validate (non-empty, ≤4000 chars)
    ├── Rate limit (per-user, in-memory)
    ├── Classify via Haiku 4.5 (small-talk / spam / question)
    │   └── small-talk → fast Haiku reply, skip context fetch
    │
    ├── Fetch context — two flows:
    │   ├── tool-use (ENABLE_TOOL_USE=true)
    │   │   Claude calls tools as needed:
    │   │   read_thread, read_channel, search_slack_history,
    │   │   search_notion, search_workforce, search_calamari,
    │   │   search_calendar, search_pipedrive, deal_status,
    │   │   create_event, create_deal_note, create_deal_activity,
    │   │   send_slack_message
    │   └── legacy — parallel fetch:
    │       Slack history + Notion + Workforce
    │
    ├── Inject knowledge files (apps/bot/knowledge/*.md)
    ├── Inject conversation history (Supabase)
    ├── Call Claude Sonnet 4.5 with system prompt + context
    │
    └── Format response (Markdown → Slack mrkdwn) and post in thread

In parallel, the bot runs:

Crawler — real-time listener that indexes all messages from channels Iwan is invited to (Supabase full-text search)
Backfill trigger — when added to a channel, asks for approval before backfilling (approvalFlow.ts)
Scheduler — 10 cron jobs (see Scheduled Jobs)
Dashboard API — Express 5 REST API for the admin dashboard

Repo structure

apps/
├── bot/                      — @iwan/bot — Slack bot + REST API
│   ├── src/
│   │   ├── index.ts          # Entry point — Socket Mode + scheduler + API
│   │   ├── api/              # Dashboard API (Express 5)
│   │   │   ├── server.ts
│   │   │   ├── routes.ts
│   │   │   └── middleware.ts # Bearer auth + role resolution
│   │   ├── handlers/
│   │   │   ├── slash.ts      # /iwan command handler
│   │   │   └── approvalFlow.ts
│   │   ├── crawler/
│   │   │   ├── listener.ts
│   │   │   ├── backfill.ts
│   │   │   ├── backfillTrigger.ts
│   │   │   └── saveMessage.ts
│   │   ├── proactive/        # Proactive engine (auto-replies in active threads)
│   │   │   ├── engine.ts
│   │   │   ├── setup.ts
│   │   │   └── ...
│   │   └── services/
│   │       ├── claude.ts, claudeHaiku.ts, claudeTools.ts
│   │       ├── tools.ts, toolExecutor.ts, authorizedExecutor.ts
│   │       ├── models.ts            # Sonnet 4.5 + Haiku 4.5
│   │       ├── cache.ts             # Redis (ioredis) — graceful degradation
│   │       ├── scheduler.ts         # Centralized cron (replaces setInterval)
│   │       ├── search.ts, memory.ts, classify.ts, validate.ts, ratelimit.ts
│   │       ├── notion.ts, channelClassification.ts (restricted DBs)
│   │       ├── workforce.ts, workforceAlerts.ts, workforceAnomaly.ts
│   │       ├── pipedrive.ts, dealResolver.ts, dealDigest.ts, dealConfig.ts
│   │       ├── channelDigest.ts, channelAnomaly.ts
│   │       ├── calamari.ts, calendar.ts
│   │       ├── knowledge.ts         # Loads knowledge/*.md into LLM prompts
│   │       ├── llm.ts, openrouter.ts, anthropicClient.ts, promptCache.ts
│   │       ├── audit.ts, errors.ts, supabase.ts
│   │       └── format.ts, users.ts, channels.ts, context.ts, membership.ts
│   ├── knowledge/            # Company context (.md, auto-loaded into prompts)
│   ├── scripts/              # backfill.js, backfillDeals.js, backfillFull.js, seed-*.sql
│   ├── tests/                # 51 Jest test files (ts-jest)
│   └── Dockerfile            # node:20-alpine multi-stage (build + runtime)
│
├── dashboard/                — @iwan/dashboard — React 19 admin SPA
│   ├── src/
│   │   ├── App.tsx, main.tsx
│   │   ├── pages/            # Health, Scheduler, Errors, Cache, Channels,
│   │   │                     # Workforce, DealDigests, Config, Login
│   │   ├── components/       # Layout, DataTable, RefreshButton, StatusBadge
│   │   └── api/              # Fetch wrappers
│   └── vite.config.ts
│
└── packages/shared/          — @iwan/shared — types + constants
    └── src/
        ├── types.ts
        ├── constants.ts      # CACHE_TTL, APP_VERSION, role/label types
        └── index.ts

Integrations

Slack

Socket Mode (@slack/bolt v4) — no public URL or webhook needed. Listens for @Iwan mentions, replies in-thread. Background crawler indexes all messages from channels Iwan is invited to, storing them in Supabase for full-text search. Supports image input (PNG/JPEG/GIF/WebP, ≤4MB, max 3 per message) via Claude vision.

Notion

@notionhq/client v5. Extracts keywords from question (Polish stop-words removed), searches workspace, fetches page content (paragraphs, headings, tables, callouts, nested blocks). Up to 3 pages per query, truncated to 1500 chars each. Supports restricted databases — channels labeled leadership see all pages; other channels are filtered (NOTION_RESTRICTED_DATABASES env).

Pipedrive CRM

Full deal-intelligence integration:

On-demand deal lookup — @Iwan status deal Acme or /iwan deal Acme
Daily digest (Mon–Fri) — auto-summarizes Slack conversations and writes them to Pipedrive deal notes with [Slack Summary] prefix
Channel-to-deal mapping — auto-resolves #sales-* channels by prefix; shared channels via LLM extraction (cached in Supabase deal_channel_mappings)
Action items — extracts next steps and creates Pipedrive activities
Backfill — node scripts/backfillDeals.js --days 7 [--dry-run] [--deal "Acme"]

Workforce Planner

Read-only integration with internal Workforce Planner (FastAPI + PostgreSQL, JWT auth). Polish date parsing — "w marcu", "Q1", "w kwietniu". Used for team allocation queries, overbooking detection, availability lookups, and two cron jobs: daily alerts (overbooking + low utilization) and weekly Monday team-allocation summary.

Calamari

Time-off / PTO lookups via Calamari API. Tool: search_calamari.

Google Calendar

Read events + create events. Service account auth (GOOGLE_SERVICE_ACCOUNT_KEY), supports multiple calendars. Tools: search_calendar, create_event. Configurable timezone (default Europe/Warsaw).

OpenRouter (LLM fallback)

When Anthropic API fails, automatically retries via OpenRouter — keeps Iwan responsive during Anthropic outages. Optional (OPENROUTER_API_KEY).

Redis (cache)

ioredis v5 with graceful degradation — no REDIS_URL = no cache, app still works. Cache-aside wrapper with TTL (withCache). Used for Pipedrive deal lookups, Notion pages, channel/user resolution.

Admin Dashboard

React 19 + Vite 6 + Tailwind 4 + TanStack Query SPA at apps/dashboard/. Talks to the bot's REST API (apps/bot/src/api/) over HTTP. Bearer-token auth with three roles:

leadership — full access, can modify channel access levels
growth — sees only open channels and growth-labeled channels
default — read-only health/config

Pages: Health · Scheduler (manual job triggers) · Errors · Cache · Channels (with access-level controls for leadership) · Workforce alerts · Deal digest history · Config (feature flags) · Login

Dashboard API endpoints

Method	Path	Auth	Description
GET	`/api/health`	none	Uptime, Redis status, job count, version
GET	`/api/scheduler/jobs`	bearer	List registered cron jobs
POST	`/api/scheduler/jobs/:name/trigger`	bearer	Manual job trigger
GET	`/api/errors`	bearer	Last 50 errors from `error_logs`
GET	`/api/cache/stats`	bearer	Redis stats (memory, key count, clients)
GET	`/api/channels`	bearer	Channel list + message counts (filtered per role)
POST	`/api/channels/:id/access`	leadership only	Set channel access level / label
GET	`/api/deals/digests`	bearer	Pipedrive digest state history
GET	`/api/workforce/alerts`	bearer	Workforce alert log
GET	`/api/audit`	bearer	Tool-use audit logs (filtered per role)
GET	`/api/config`	bearer	Safe feature flags (no secrets)

Disabled by default. Enable with ENABLE_DASHBOARD_API=true and set DASHBOARD_API_TOKEN (+ optional DASHBOARD_TOKEN_LEADERSHIP / DASHBOARD_TOKEN_GROWTH for role separation).

Slash Commands

All commands are currently Polish-only. English aliases are the v0.6 priority.

Command	What it does
`/iwan szukaj <fraza>`	Full-text search in Slack message history (Supabase RPC)
`/iwan notion <fraza>`	Search Notion pages by keyword (with restricted DB filtering)
`/iwan team <nazwa>`	Show team members with utilization % (e.g. `team Backend`)
`/iwan kto-wolny [miesiąc]`	List people with <30% utilization
`/iwan overbooking`	List people with >100% utilization (next 2 months)
`/iwan projekty`	List active projects
`/iwan deal <nazwa>`	Show Pipedrive deal status (CRM data + recent notes)
`/iwan deals [pipeline_id]`	List active deals from configured pipelines
`/iwan status`	Bot uptime, memory, Node version

Restrict commands to specific channels with SLACK_ALLOWED_CHANNELS (comma-separated channel IDs).

Scheduled Jobs

Centralized in services/scheduler.ts (node-cron v4, timezone Europe/Warsaw). Jobs register conditionally based on env vars.

Job	Schedule	Condition
`health-check`	every 5 min	always
`deal-digest`	Mon–Fri at `DEAL_DIGEST_HOUR` (default 7)	`PIPEDRIVE_API_TOKEN`
`deal-inactive-check`	Mon–Fri at 9:00	`PIPEDRIVE_API_TOKEN`
`workforce-alerts`	daily at 8:00	`WP_ALERT_CHANNEL` + `WP_API_URL`
`workforce-alerts-cleanup`	daily at 3:00	`WP_ALERT_CHANNEL`
`workforce-weekly-summary`	Mondays at `WP_SUMMARY_HOUR` (default 8)	`WP_SUMMARY_CHANNEL` + `WP_API_URL`
`workforce-anomaly`	Mon–Fri at 9:00	`WP_ALERT_CHANNEL` + `WP_API_URL`
`channel-digest`	Mon–Fri at `CHANNEL_DIGEST_HOUR` (default 8)	`CHANNEL_DIGEST_ENABLED=true`
`channel-anomaly`	every 30 min	`CHANNEL_ANOMALY_ENABLED=true`
`proactive-cleanup`	every hour	`ENABLE_PROACTIVE=true`

Trigger any job manually via dashboard or POST /api/scheduler/jobs/:name/trigger.

Getting Started

Prerequisites

Node.js 20.x
pnpm 9 (corepack enable && corepack prepare pnpm@9 --activate)
Slack workspace with Bot Token + App Token (Socket Mode enabled)
Supabase project (tables: slack_messages, conversations, error_logs, audit_logs, deal_channel_mappings, deal_digest_state, channel_access_levels)
Anthropic API key
(optional) Redis instance for caching (REDIS_URL)
(optional) Notion / Pipedrive / Workforce Planner / Calamari / Google Calendar

Installation

git clone <your-repo-url> iwan
cd iwan
pnpm install --frozen-lockfile

cp .env.example .env            # then fill in your credentials
psql ... -f apps/bot/scripts/seed-deal-tables.sql
psql ... -f apps/bot/scripts/seed-access-control.sql
psql ... -f apps/bot/scripts/seed-company-context.sql

pnpm turbo typecheck            # full typecheck across monorepo
pnpm turbo test                 # 51 Jest test files (apps/bot)
pnpm turbo build                # build shared + bot + dashboard

# Run bot
node apps/bot/dist/src/index.js

# Run dashboard (dev)
pnpm --filter @iwan/dashboard dev

# Run bot (dev with hot reload)
pnpm --filter @iwan/bot dev      # uses tsx

Docker

docker compose up               # see docker-compose.yml
# or
docker build -f apps/bot/Dockerfile -t iwan .
docker run --env-file .env iwan

The Dockerfile is multi-stage (node:20-alpine): builds shared + bot + dashboard with pnpm turbo build, then ships only dist/ and prod deps.

Environment Variables

Required

Variable	Description
`SLACK_BOT_TOKEN`	Slack bot token (`xoxb-...`)
`SLACK_APP_TOKEN`	Slack app-level token (`xapp-...`) for Socket Mode
`ANTHROPIC_API_KEY`	Anthropic API key for Claude Sonnet 4.5 + Haiku 4.5
`SUPABASE_URL`	Supabase project URL
`SUPABASE_KEY`	Supabase anon key

Slack — optional

Variable	Description
`SLACK_ADMIN_USER_ID`	Slack user ID with admin privileges (approves backfills)
`SLACK_ALLOWED_CHANNELS`	Comma-separated channel IDs — restrict slash commands

Tool-use mode — optional

Variable	Default	Description
`ENABLE_TOOL_USE`	`false`	When `true`, Claude decides which tools to call. When `false`, parallel-fetch legacy mode.

Cache — optional

Variable	Description
`REDIS_URL`	e.g. `redis://localhost:6379`. Without this, all cache reads return null and writes are skipped.

Notion — optional

Variable	Description
`NOTION_TOKEN`	Notion integration token (`secret_...`)
`NOTION_RESTRICTED_DATABASES`	Comma-separated DB IDs visible only to leadership channels

Pipedrive CRM — optional

Variable	Default	Description
`PIPEDRIVE_API_TOKEN`	—	Without this, all CRM features and `deal-*` jobs are disabled
`PIPEDRIVE_DOMAIN`	—	Subdomain (e.g. `your-company`)
`PIPEDRIVE_ACTIVE_PIPELINES`	`1`	Comma-separated pipeline IDs to monitor
`DEAL_DIGEST_CHANNEL`	—	Slack channel for digest status messages
`DEAL_DIGEST_HOUR`	`7`	Hour to run daily digest (0–23)
`DEAL_SALES_PREFIX`	`sales-`	Prefix for auto-discovered deal channels
`DEAL_MONITORED_CHANNELS`	`deals`	Comma-separated shared channel names
`DEAL_MIN_MESSAGES`	`3`	Min messages before a thread gets summarized
`DEAL_NOTE_PREFIX`	`[Slack Summary]`	Prefix on Pipedrive notes
`DEAL_LANGUAGE`	`pl`	Summary language

Workforce Planner — optional

Variable	Default	Description
`WP_API_URL`	—	Workforce Planner API base URL
`WP_EMAIL`	—	Login email (JWT auth)
`WP_PASSWORD`	—	Login password
`WP_ALERT_CHANNEL`	—	Channel for overbooking + anomaly alerts
`WP_ALERT_INTERVAL_HOURS`	`24`	How often to check
`WP_LOW_UTIL_THRESHOLD`	`20`	Low utilization alert threshold (%)
`WP_SUMMARY_CHANNEL`	—	Channel for Monday weekly summary
`WP_SUMMARY_HOUR`	`8`	Hour to post Monday summary (0–23)
`WORKFORCE_ANOMALY_ALLOC_DROP_PCT`	`30`	Allocation-drop alert threshold
`WORKFORCE_ANOMALY_ALLOC_SPIKE_PCT`	`50`	Allocation-spike alert threshold

Calamari + Google Calendar — optional

Variable	Default	Description
`CALAMARI_URL`	—	e.g. `https://yourcompany.calamari.io`
`CALAMARI_API_KEY`	—	API key
`GOOGLE_SERVICE_ACCOUNT_KEY`	—	JSON service account credentials
`GOOGLE_CALENDAR_IDS`	—	Comma-separated calendar IDs
`GOOGLE_CALENDAR_TIMEZONE`	`Europe/Warsaw`	IANA timezone

Proactive engine — optional

Variable	Default	Description
`ENABLE_PROACTIVE`	`false`	Enable proactive auto-replies in active threads
`PROACTIVE_CHANNELS`	`general,team`	Channels to monitor
`PROACTIVE_THREAD_THRESHOLD`	`5`	Min thread messages before considering reply
`PROACTIVE_CHANNEL_MESSAGE_INTERVAL`	`15`	Min message gap before channel-level reply
`PROACTIVE_CONFIDENCE_THRESHOLD`	`0.7`	Min Claude confidence to send
`PROACTIVE_GLOBAL_MAX_PER_HOUR`	`10`	Global rate limit
`PROACTIVE_THREAD_COOLDOWN_MINUTES`	`60`	Per-thread cooldown
`PROACTIVE_CHANNEL_COOLDOWN_MINUTES`	`30`	Per-channel cooldown

Channel digest + anomaly — optional

Variable	Default	Description
`CHANNEL_DIGEST_ENABLED`	`false`	Daily channel summary
`CHANNEL_DIGEST_HOUR`	`8`	Run hour
`CHANNEL_DIGEST_CHANNEL`	—	Where to post digest
`CHANNEL_DIGEST_CHANNELS`	`general,team`	Channels to summarize
`CHANNEL_ANOMALY_ENABLED`	`false`	Detect message-volume spikes
`CHANNEL_ANOMALY_CHANNEL`	—	Where to post anomaly alerts
`CHANNEL_ANOMALY_CHANNELS`	`general,team`	Channels to monitor
`CHANNEL_ANOMALY_SPIKE_MULTIPLIER`	`3`	Multiplier over baseline to trigger

LLM fallback — optional

Variable	Description
`OPENROUTER_API_KEY`	OpenRouter key — fallback when Anthropic API fails

Dashboard API — optional

Variable	Default	Description
`ENABLE_DASHBOARD_API`	`false`	Enable REST API server
`DASHBOARD_API_PORT`	`3100`	Listen port
`DASHBOARD_API_TOKEN`	—	Default bearer token (read-only)
`DASHBOARD_TOKEN_LEADERSHIP`	—	Leadership-role bearer token
`DASHBOARD_TOKEN_GROWTH`	—	Growth-role bearer token

Tech Stack

Component	Tool	Version
Runtime	Node.js	20.x
Language	TypeScript (strict, NodeNext)	5.9
Monorepo	pnpm workspaces + Turborepo	pnpm 9.15, turbo 2.5
Slack	`@slack/bolt` (Socket Mode)	4.6
AI (answers)	Claude Sonnet 4.5 via `@anthropic-ai/sdk`	`claude-sonnet-4-5-20250929`
AI (classification + small-talk)	Claude Haiku 4.5	`claude-haiku-4-5-20251001`
LLM fallback	OpenRouter	—
API	Express (Dashboard REST API)	5.x
Database	Supabase (PostgreSQL + full-text search)	2.97
Cache	Redis via `ioredis` (graceful degradation)	5.10
Scheduler	`node-cron` (timezone Europe/Warsaw)	4.2
Knowledge base	Notion API (`@notionhq/client`)	5.9
Workforce data	Workforce Planner (FastAPI, JWT)	—
PTO	Calamari API	—
Calendar	`googleapis`	171
Dashboard	React 19 + Vite 6 + Tailwind 4 + TanStack Query	—
Tests	Jest + ts-jest (51 test files in `apps/bot/tests/`)	29
CI	GitHub Actions (pnpm + turbo typecheck/test/build)	—
Container	Docker (`node:20-alpine`, multi-stage)	—
Hosting	Railway	—

Contributing

Iwan is MIT-licensed. Contributions welcome — bug fixes, new features, docs, tests.

How to Contribute

Fork the repo and clone locally
pnpm install --frozen-lockfile
Create a branch: git checkout -b feat/my-feature
Write code following the conventions below
Add tests in apps/bot/tests/
Run pnpm turbo typecheck && pnpm turbo test && pnpm turbo build — all green
Open a PR

Conventions

Rule	Detail
One function = one task	`searchWorkforce()` searches. `buildContextFromWorkforce()` formats. They don't do both.
Max 30 lines per function	If it's longer, split it.
Comments in Polish above functions	Polish-origin project. `// Pobierz timeline alokacji`
English variable names	`const employees = ...`, `const startDate = ...`
Don't overwrite, add new files	Especially in services/ — new file > big edit to existing
Graceful degradation	Service functions return `[]`, `''`, or `null` on error — never throw
TypeScript strict, no `any` (or annotated)	`// eslint-disable-next-line @typescript-eslint/no-explicit-any` if unavoidable

Good First Contributions

English command aliases — /iwan search, /iwan who-free, /iwan projects. Start with apps/bot/src/handlers/slash.ts.
English keyword routing — shouldQueryWorkforce() in apps/bot/src/services/workforce.ts only detects Polish phrases.
Configurable response language — currently hardcoded Polish in system prompts.
English date parsing — Workforce date helpers parse "w marcu", "Q1" etc., not "next month" / "in March".

Roadmap

v0.6 — English Commands & Enhanced Workforce

English command aliases — /iwan search, /iwan who-free, /iwan projects
English keyword routing for Workforce queries
Configurable response language (EN/PL)
/iwan person <name> — individual person lookup
/iwan projekty — show assigned people per project
Smarter date parsing — "next week", "next month"

v0.7 — Semantic Search

Voyage AI embeddings to replace keyword-based full-text search
pgvector in Supabase for vector similarity
Cross-source ranking — prioritize most relevant results
Thread-aware context — include parent thread messages

v0.8 — Proactive Intelligence

Daily channel summary — "what happened yesterday"
Capacity planning — "do we have people for a new project in Q2?"
Utilization trend tracking over time
Pipedrive deal-health scoring

v1.0 — Production Ready

Multi-Workspace support (phase 7)
Nango for managed external integrations
Configurable system prompt per workspace
Workspace-level rate limiting
Persistent JWT token storage (replace in-memory)
Two-way Pipedrive sync (Pipedrive → Slack notifications)

v2.0 — Autonomous Agent

E2B sandbox for code execution
Write-back to Workforce Planner (create assignments)
Jira / Linear integration
Automated resource suggestions based on skills + availability

Directional only — priorities shift based on what users need.

Iwan is built by Momentum and the open source community.

If you find it useful, a star helps others discover it.

MIT License

Name		Name	Last commit message	Last commit date
Latest commit History 96 Commits
.claude		.claude
.github/workflows		.github/workflows
apps		apps
docs		docs
packages/shared		packages/shared
.env.example		.env.example
.gitignore		.gitignore
.npmrc		.npmrc
.turborc		.turborc
CHANGELOG.md		CHANGELOG.md
CLAUDE.md		CLAUDE.md
LICENSE		LICENSE
README.md		README.md
docker-compose.yml		docker-compose.yml
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
pnpm-workspace.yaml		pnpm-workspace.yaml
railway.toml		railway.toml
turbo.json		turbo.json

Folders and files

Latest commit

History

Repository files navigation

Iwan

Open Source AI Agent for Slack — with Admin Dashboard

What Iwan Does

Current State (v0.5)

How It Works

Integrations

Slack

Notion

Pipedrive CRM

Workforce Planner

Calamari

Google Calendar

OpenRouter (LLM fallback)

Redis (cache)

Admin Dashboard

Dashboard API endpoints

Slash Commands

Scheduled Jobs

Getting Started

Prerequisites

Installation

Docker

Environment Variables

Tech Stack

Contributing

How to Contribute

Conventions

Good First Contributions

Roadmap

v0.6 — English Commands & Enhanced Workforce

v0.7 — Semantic Search

v0.8 — Proactive Intelligence

v1.0 — Production Ready

v2.0 — Autonomous Agent

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages