slackhive 0.1.23 → 0.1.25

package/README.md

@@ -4,7 +4,8 @@
 
 # SlackHive
 
-### Build, deploy, and orchestrate teams of Claude Code AI agents on Slack
+### Build your AI-first company on Slack
+#### Deploy and orchestrate Claude Code agents as your teammates
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
 [![npm](https://img.shields.io/npm/v/slackhive?color=cb3837&logo=npm&logoColor=white)](https://www.npmjs.com/package/slackhive)
@@ -12,10 +13,10 @@
 [![Node.js](https://img.shields.io/badge/Node.js-≥20-339933?logo=node.js&logoColor=white)](https://nodejs.org)
 [![TypeScript](https://img.shields.io/badge/TypeScript-5.x-3178c6?logo=typescript&logoColor=white)](https://www.typescriptlang.org)
 [![Docker](https://img.shields.io/badge/Docker-Compose-2496ed?logo=docker&logoColor=white)](https://docs.docker.com/compose)
-[![Claude Code SDK](https://img.shields.io/badge/Claude_Code-SDK-cc785c?logo=anthropic&logoColor=white)](https://docs.anthropic.com/en/agent-sdk)
-[![Slack](https://img.shields.io/badge/Slack-Bolt-4A154B?logo=slack&logoColor=white)](https://api.slack.com/bolt)
+[![Documentation](https://img.shields.io/badge/docs-slackhive.mintlify.app-D97757?logo=gitbook&logoColor=white)](https://slackhive.mintlify.app)
+[![Security Audit](https://github.com/pelago-labs/slackhive/actions/workflows/audit.yml/badge.svg)](https://github.com/pelago-labs/slackhive/actions/workflows/audit.yml)
 
-[Quick Start](#-quick-start) · [Features](#-features) · [Architecture](#-architecture) · [Documentation](#-creating-your-first-agent) · [Contributing](#-contributing)
+[**Documentation**](https://slackhive.mintlify.app) · [Quick Start](#-quick-start) · [Features](#-features) · [Architecture](#-architecture) · [Contributing](#-contributing)
 
 </div>
 
@@ -23,127 +24,37 @@
 
 ## Why SlackHive?
 
-Most AI agent frameworks focus on a single agent doing a single task. But real teams have **specialists** — a data analyst who speaks SQL, a writer who crafts announcements, an engineer who reviews code. What if your AI team worked the same way?
+Your Slack workspace is where your team already lives. Every question, decision, and escalation happens there. SlackHive makes that workspace a mix of **people and agents** — side by side, in the same channels, in the same threads.
 
-**SlackHive** was born from a simple observation: the most powerful AI setup isn't one omniscient agent — it's a **team of specialists** that learn and improve from every interaction.
-
-Inspired by how engineering teams actually collaborate in Slack, we built a platform where:
-
-- **Each agent is a Slack bot** with its own identity, skills, and memory
-- **A Boss Agent** knows the entire team and delegates work by @mentioning the right specialist
-- **Every agent learns** — memories from conversations are persisted and loaded on the next start
-- **Everything is configurable** from a clean web UI — no code changes needed to add agents, assign tools, or edit behavior
-
-Whether you're building an internal AI ops team, a customer support squad, or a research group — SlackHive gives you the infrastructure to make it happen.
-
-<details>
-<summary><b>See it in action</b></summary>
+These aren't chatbots you switch to. They're colleagues you @mention. Each agent connects to the tools your team already uses — Notion, Jira, GitHub, Figma, your database, your analytics stack. **Anyone on the team can create one.** No engineers, no platform team — if you can describe what you need, you can deploy it in minutes.
 
 ```
-User:     @boss can you analyze last week's conversion funnel?
-Boss:     That's right up @data-analyst's alley. Let me loop them in 👇
-          @data-analyst — user wants conversion funnel analysis for last week.
-DataBot:  [reads full thread context, runs Redshift query via MCP]
-          Here are the results: conversions were up 12% week-over-week...
+CEO:        @data-analyst revenue is down 8% this week, can you dig in?
+DataBot:    [queries Redshift across 6 dimensions]
+            Found it — enterprise churn spiked Tuesday after the pricing change.
+            3 accounts, $42k ARR at risk.
+
+Engineer:   @devops the checkout service is throwing 500s
+DevOps:     [reads logs, identifies root cause, opens PR]
+            Memory leak in the payment processor pool. PR #847 is up with the fix.
+
+PM:         @designer mock up a simpler onboarding flow
+Designer:   [creates Figma frames via MCP]
+            Done — 3 variants in Figma. Which direction do you want to take?
 ```
 
-The Boss reads the message, checks its team registry, and delegates to the right specialist. The specialist picks up the **full Slack thread** as context — nothing is lost in the handoff.
-
-</details>
-
----
-
-## ✨ Features
-
-### Core Platform
-
-| Feature | Description |
-|---------|-------------|
-| 👑 **Boss Agent** | Orchestrator bot that knows every specialist and delegates by @mention in threads |
-| 🧠 **Agent Memory** | Agents learn from every conversation — memories auto-synced to Postgres |
-| 🔌 **MCP Server Catalog** | Add tool servers once, assign to any agent — Redshift, GitHub, custom APIs |
-| 🧵 **Thread Context** | Tagged agents fetch full thread history — zero context loss in handoffs |
-| 💾 **Session Persistence** | Slack thread ↔ Claude session mapping survives restarts |
-| 🔁 **Hot Reload** | Edit anything in the UI → agent picks up changes in seconds via Redis pub/sub |
-
-### Web UI
-
-| Feature | Description |
-|---------|-------------|
-| 🧙 **Onboarding Wizard** | Guided flow: identity → Slack app → tokens → MCPs & skills (skipped for boss) → review |
-| 📝 **Skill Editor** | In-browser editor for agent markdown skills with file tree and categories |
-| 🔐 **Tool Permissions** | Per-agent allowlist/denylist for Claude Code SDK tools |
-| 📊 **Live Logs** | SSE-streamed Docker log output per agent with level filters and search |
-| ⚙️ **Settings** | Configurable branding (app name, logo, tagline), dashboard title, user management |
-| 🧠 **Memory Viewer** | Browse, inspect, and delete agent memories grouped by type |
-| 📄 **CLAUDE.md Viewer** | Read and edit the compiled system prompt sent to each agent |
-| 📐 **Collapsible Sidebar** | Clean sidebar with live agent roster, status dots, and collapse toggle |
-| 📱 **Responsive Design** | Mobile-friendly layout with hamburger menu, overlay sidebar, fluid grids |
-| 🔒 **Auth & RBAC** | Login page, superadmin via env vars, 4 roles (superadmin/admin/editor/viewer) |
-| 👥 **User Management** | Create users with admin, editor, or viewer roles from Settings |
-| 🏢 **Agent Hierarchy** | Multi-boss support — agents can report to multiple bosses, each boss manages its own team |
-| ⏰ **Scheduled Jobs** | Cron-based recurring tasks executed by the boss agent, with run history |
-
-### Agent Capabilities
-
-- **Slack Block Kit formatting** — markdown tables rendered as native Slack table blocks, headings, code blocks
-- **Streaming responses** — tool use labels, progress indicators, and rich formatted output
-- **MCP tool integration** — stdio, SSE, and HTTP transports supported
-- **Customizable personas** — each agent has its own personality and behavior
-- **Skill system** — modular markdown files organized by category with sort ordering
-- **Auto-generated boss registry** — each boss gets a team roster compiled from agents that report to it
-- **Memory system injected into CLAUDE.md** — agents know how to write and organize memories
-- **Multi-boss hierarchy** — `reports_to` is a UUID array; an agent can report to multiple bosses
-
----
-
-## 🏗 Architecture
+Tag a specialist directly when you know who to ask. Or tag `@boss` when you're not sure — Boss finds the right specialist, delegates, and summarizes the result:
 
 ```
-┌─────────────────────────────────────────────────────────────┐
-│  Slack Workspace                                            │
-│  @boss  @data-bot  @writer  @engineer  ...                  │
-└────────────────────────────┬────────────────────────────────┘
-                             │ Socket Mode (Bolt)
-┌────────────────────────────▼────────────────────────────────┐
-│  Docker Compose                                             │
-│                                                             │
-│  ┌─────────────────┐  publish events  ┌─────────────────┐   │
-│  │  Web UI         │ ──────────────►  │  Redis 7        │   │
-│  │  Next.js 15     │                  │  pub/sub        │   │
-│  │  :3000          │                  └────────┬────────┘   │
-│  │                 │                           │ subscribe  │
-│  │  • Dashboard    │  read/write       ┌───────▼────────┐   │
-│  │  • Agent config │ ◄───────────────► │  Runner        │   │
-│  │  • Skill editor │                   │                │   │
-│  │  • MCP catalog  │                   │  AgentRunner   │   │
-│  │  • Memory viewer│                   │  ├─ Boss       │   │
-│  │  • Live logs    │                   │  ├─ DataBot    │   │
-│  │  • Settings     │                   │  ├─ Writer     │   │
-│  └─────────────────┘                   │  └─ ...        │   │
-│          │                             └───────┬────────┘   │
-│          │ read/write                          │            │
-│          ▼                                     ▼            │
-│  ┌─────────────────────────────────────────────────────┐    │
-│  │  PostgreSQL 16                                      │    │
-│  │                                                     │    │
-│  │  agents · skills · memories · permissions           │    │
-│  │  mcp_servers · agent_mcps · sessions                │    │
-│  │  settings · users · scheduled_jobs · job_runs       │    │
-│  └─────────────────────────────────────────────────────┘    │
-└─────────────────────────────────────────────────────────────┘
+You:        @boss can you analyze last week's conversion funnel?
+Boss:       That's right up @data-analyst's alley 👇
+            @data-analyst — conversion funnel analysis for last week.
+            When you're done, please tag @boss.
+DataBot:    Conversions up 12% WoW, checkout completion jumped 3×. @boss — done!
+Boss:       Conversions are up 12% WoW. The win was checkout — 3× completion rate.
+            Want me to pull a channel or cohort breakdown?
 ```
 
-**How it flows:**
-
-1. **User** messages an agent (or `@boss`) in Slack
-2. **Runner** receives the event via Bolt Socket Mode
-3. **Claude Code SDK** processes the message with the agent's compiled `CLAUDE.md`
-4. Agent may use **MCP tools** (Redshift queries, GitHub API, etc.) during processing
-5. **Response** is formatted as Slack Block Kit and posted to the thread
-6. **Memory files** written during the session are detected by `MemoryWatcher` and synced to Postgres
-7. On next conversation, the agent starts with all accumulated **learned knowledge**
-
 ---
 
 ## 🚀 Quick Start
@@ -155,63 +66,10 @@ npm install -g slackhive
 slackhive init
 ```
 
-The CLI will:
-1. Check prerequisites (Docker, Docker Compose, Git)
-2. Clone the repository
-3. Walk you through configuration (API key, admin credentials)
-4. Start all services automatically
-
-### Option B: Manual setup
-
-#### Prerequisites
-
-- [Docker](https://docs.docker.com/get-docker/) + [Docker Compose](https://docs.docker.com/compose/)
-- An [Anthropic API key](https://console.anthropic.com/) (`ANTHROPIC_API_KEY`)
-
-#### 1. Clone & configure
-
-```bash
-git clone https://github.com/amansrivastava17/slackhive.git
-cd slackhive
-cp .env.example .env
-```
-
-Edit `.env` with your Anthropic API key and credentials:
-
-```env
-ANTHROPIC_API_KEY=sk-ant-...
-ADMIN_USERNAME=admin
-ADMIN_PASSWORD=changeme
-POSTGRES_PASSWORD=slackhive
-```
-
-#### 2. Start everything
-
-```bash
-docker compose up -d --build
-```
-
-This launches all four services:
-
-| Service | Port | Description |
-|---------|------|-------------|
-| **Web UI** | `localhost:3001` | Dashboard and agent management |
-| **Runner** | — | Manages all Slack bot connections |
-| **PostgreSQL** | `localhost:5432` | Persistent storage |
-| **Redis** | `localhost:6379` | Event pub/sub for hot reload |
-
-#### 3. Open the dashboard
-
-```
-http://localhost:3001
-```
-
-Login with your admin credentials and create your first agent.
+The CLI will check prerequisites (Docker, Git), clone the repo, walk you through configuration, and start all services automatically. Open `http://localhost:3001` and create your first agent.
 
 ### CLI Commands
 
-After installing with `npm install -g slackhive`:
-
 | Command | Description |
 |---------|-------------|
 | `slackhive init` | Clone, configure, and start SlackHive |
@@ -221,362 +79,196 @@ After installing with `npm install -g slackhive`:
 | `slackhive logs` | Tail runner logs |
 | `slackhive update` | Pull latest changes and rebuild |
 
----
-
-## 🔑 Claude Code Authentication
-
-SlackHive supports two authentication modes for the Claude Code SDK. Choose the one that fits your setup.
-
-### Option 1: API Key (pay-per-use)
-
-Best for: teams, production, predictable billing.
-
-Set your Anthropic API key in `.env`:
-
-```env
-ANTHROPIC_API_KEY=sk-ant-api03-...
-```
-
-That's it. Every agent will use this key. You're billed per token via the [Anthropic API](https://console.anthropic.com/).
-
-### Option 2: Claude Code Subscription (Max plan)
-
-Best for: individual developers, Claude Pro/Max subscribers ($100–$200/month unlimited).
-
-If you have a Claude Max subscription with Claude Code access:
+### Option B: Manual setup
 
-**Step 1 — Login on the host machine:**
+**Prerequisites:** [Docker](https://docs.docker.com/get-docker/) + [Docker Compose](https://docs.docker.com/compose/), Node.js ≥ 20
 
 ```bash
-claude login
-```
-
-This opens a browser for OAuth and saves credentials to `~/.claude/`.
-
-**Step 2 — Mount credentials into the runner container:**
-
-The `docker-compose.yml` runner service needs access to your host's Claude credentials. Add these volume mounts if not already present:
-
-```yaml
-runner:
-  volumes:
-    - ~/.claude:/root/.claude          # Auth credentials
-    - /tmp/agents:/tmp/agents          # Agent working dirs
+git clone https://github.com/pelago-labs/slackhive.git
+cd slackhive
+cp .env.example .env
 ```
 
-**Step 3 — Remove the API key (important):**
-
-Make sure `ANTHROPIC_API_KEY` is **not** set in `.env`. When no API key is present, the SDK falls back to the subscription credentials from `~/.claude/`.
+Edit `.env` with your credentials:
 
 ```env
-# ANTHROPIC_API_KEY=          ← comment out or remove
+ANTHROPIC_API_KEY=sk-ant-...        # or use Claude Pro/Max subscription
+ADMIN_USERNAME=admin
+ADMIN_PASSWORD=changeme
+POSTGRES_PASSWORD=slackhive
+ENV_SECRET_KEY=                     # generate: openssl rand -hex 32
 ```
 
-**Step 4 — Restart:**
-
 ```bash
-slackhive update
-# or: docker compose up -d --build runner
+docker compose up -d --build
 ```
 
-### Which should I use?
-
-| | API Key | Subscription |
-|---|---------|-------------|
-| **Billing** | Per-token (pay what you use) | Flat monthly ($100/$200) |
-| **Setup** | Just paste the key | Run `claude login` on host |
-| **Best for** | Teams, CI/CD, production | Solo devs, prototyping |
-| **Rate limits** | API tier limits | Subscription fair-use limits |
-| **Multiple agents** | All share one key | All share one subscription |
+Open `http://localhost:3001`, log in, and create your first agent.
 
-> **Note:** If both `ANTHROPIC_API_KEY` and `~/.claude` credentials are present, the API key takes precedence.
+> Full setup guide → [slackhive.mintlify.app/quickstart](https://slackhive.mintlify.app/quickstart)
 
 ---
 
-## 🤖 Creating Your First Agent
-
-Click **New Agent** from the dashboard and follow the wizard:
-
-### Step 1 — Identity
-Set the agent's name, slug (e.g., `data-bot`), persona, and description. Toggle **Boss** if this agent should orchestrate others — boss agents skip the MCPs & Skills step since their `CLAUDE.md` is auto-generated. For specialist agents, select which boss(es) they report to.
-
-### Step 2 — Slack App Setup
-The platform generates a `slack-app-manifest.json`. Create a Slack app from this manifest:
-
-1. Go to [api.slack.com/apps](https://api.slack.com/apps) → **Create New App** → **From a manifest**
-2. Paste the generated JSON
-3. **Install to Workspace** → copy the **Bot Token** (`xoxb-...`)
-4. **Socket Mode** → Enable → generate the **App-Level Token** (`xapp-...`)
-5. **Basic Information** → copy the **Signing Secret**
-6. Paste all three back in the wizard
-
-### Step 3 — Permissions
-Configure which Claude Code SDK tools the agent can use. Quick-add buttons for common tools like `Read`, `Write`, `Bash`, `WebFetch`.
-
-### Step 4 — MCPs
-Select MCP servers from the platform catalog to give your agent access to external tools (databases, APIs, etc.).
-
-### Step 5 — Skills
-Choose a starter template or start blank. Skills are modular markdown files that compose the agent's `CLAUDE.md` system prompt.
-
-The agent starts automatically and connects to Slack.
-
----
-
-## 👑 Boss Agents
-
-Create one or more agents with the **Boss** toggle enabled. Each boss gets its own `CLAUDE.md` team registry listing the agents that report to it:
-
-```markdown
-## Your Team
-
-- **DataBot** (<@U12345678>) — Data warehouse NLQ, Redshift queries, business metrics
-- **Writer** (<@U87654321>) — Content generation, Slack summaries, announcements
-```
-
-The registry **auto-regenerates** for every boss whenever you add, update, or delete an agent — no manual maintenance needed.
-
-You can have **multiple boss agents** for different domains or teams. When creating a specialist agent, check the bosses it should report to — it will appear in each selected boss's registry. A specialist can report to more than one boss.
+## ✨ Features
 
-```
-User:     @boss help me analyze last week's bookings
-Boss:     I'll get @data-bot on this 👇
-          @data-bot — user wants booking analysis for last week
-DataBot:  [reads full thread, runs Redshift query via MCP]
-          Bookings were up 12% to 4,320...
-```
+### 🤖 Real AI Agents — Not Chatbots
+
+Every agent is a full **Claude Code** agent — with tools, memory, identity, and instructions. When you @mention one in Slack, you're running a real AI agent that can use tools, take action, and get smarter over time.
+
+| | |
+|---|---|
+| 🧠 **Persistent Memory** | Agents write memories during conversations — feedback, user context, project state. Synced to Postgres, injected on next start. They don't forget. |
+| 🔌 **MCP Tool Integration** | Connect any MCP server (Redshift, GitHub, Notion, Figma, custom APIs) — stdio, SSE, or HTTP transports. |
+| 📝 **Inline TypeScript MCPs** | Paste TypeScript source directly into the UI — no deployment needed. The runner compiles and executes it. |
+| 🧵 **Full Thread Context** | Agents fetch the entire Slack thread on every invocation — zero context lost in handoffs. |
+| 💾 **Session Continuity** | Slack thread ↔ Claude session mapping survives restarts. Pick up exactly where you left off. |
+| 🔐 **Encrypted Secret Store** | API keys encrypted at rest (AES-256). MCPs reference secrets by name — raw values never touch the API or UI. |
+| 🔁 **Hot Reload** | Edit instructions, skills, or tools and the agent picks up changes within seconds. No restart needed. |
+
+### 👑 Boss + Specialist Hierarchy
+
+| | |
+|---|---|
+| 👑 **Boss Orchestration** | Boss reads your message, finds the right specialist, delegates by @mention in the same thread, and summarizes the result. |
+| 🏢 **Multi-Boss Support** | Run multiple Boss agents for different domains (engineering, data, support). Specialists can report to more than one boss. |
+| 📋 **Auto-Generated Registries** | Every Boss gets a live team roster auto-regenerated whenever the team changes. No manual maintenance. |
+| 🛠 **Skills** | Markdown files deployed as Claude Code slash commands. Give agents SQL rules, writing guidelines, or domain playbooks. |
+| ⏰ **Scheduled Jobs** | Cron-based recurring tasks — daily reports, weekly digests, monitoring alerts — posted to any Slack channel or DM. |
+
+### ⚙️ Platform
+
+| | |
+|---|---|
+| 🧙 **Onboarding Wizard** | 5-step guided setup: identity → Slack app → credentials → tools & skills → review. |
+| 🕓 **Version Control** | Every save auto-snapshots the full agent state. Browse history with line-level diffs, restore any version in one click. |
+| 🔒 **Auth & RBAC** | 4 roles (superadmin / admin / editor / viewer), HMAC-signed sessions, per-agent write access grants. No external auth provider needed. |
+| 🚦 **Channel Restrictions** | Lock agents to specific Slack channels. Bot auto-leaves uninvited channels with a notice. |
+| 📊 **Live Logs** | SSE-streamed log output per agent — with level filters and search. |
+| 🧠 **Memory Viewer** | Browse, inspect, and delete agent memories by type — feedback, user, project, reference. |
 
 ---
 
-## 🧠 How Agents Learn
-
-Every conversation is an opportunity for the agent to learn. This is the **primary design goal** of the platform.
+## 🏗 Architecture
 
 ```
-Conversation
-  └─► Claude writes .claude/memory/feedback_xyz.md
-        └─► MemoryWatcher detects change (fs.watch)
-              └─► Parses YAML frontmatter (name, type, description)
-                    └─► Upserts into memories table (Postgres)
-                          └─► Included in CLAUDE.md on next start
+Slack Workspace (@boss, @data-bot, @writer, ...)
+        │ Socket Mode (Bolt)
+        ▼
+┌──────────────────────────────────────────────────┐
+│  Docker Compose                                  │
+│                                                  │
+│  Web (Next.js) ──── Redis ────► Runner           │
+│       │                          │               │
+│       └──────── PostgreSQL ──────┘               │
+└──────────────────────────────────────────────────┘
 ```
 
-Memory types follow [Claude Code memory conventions](https://docs.anthropic.com/en/claude-code/memory):
-
-| Type | Purpose | Example |
-|------|---------|---------|
-| `feedback` | Behavioral corrections and validated approaches | "Don't mock the database in integration tests" |
-| `user` | Information about people the agent works with | "Kai is the data team lead, prefers concise answers" |
-| `project` | Ongoing work context, goals, deadlines | "Merge freeze starts March 5 for mobile release" |
-| `reference` | Pointers to external systems and resources | "Pipeline bugs tracked in Linear project INGEST" |
-
-View and manage all memories from **Agents → [name] → Memory**.
-
----
-
-## ⏰ Scheduled Jobs
-
-Scheduled jobs let the boss agent run recurring tasks on a cron schedule and post results to Slack.
-
-### How it works
-
-1. Create a job from the **Jobs** page in the web UI
-2. Set a **prompt** (what to tell the boss), **schedule** (cron expression), and **target** (channel or DM)
-3. The runner's `JobScheduler` fires on schedule and sends the prompt to the boss agent
-4. Boss processes it like any normal message — may delegate to specialists, run MCP tools, etc.
-5. Result is posted to the target Slack channel or DM
-6. Run history (status, output, duration) is tracked and visible in the UI
-
-### Example
-
-| Field | Value |
-|-------|-------|
-| **Name** | Daily Booking Report |
-| **Prompt** | Generate a summary of yesterday's bookings with key metrics |
-| **Schedule** | `0 8 * * *` (daily at 8:00 AM) |
-| **Target** | `#analytics` channel |
-
-The UI includes schedule presets (hourly, daily, weekdays, weekly) and shows cron expressions in human-readable form.
-
-### Job run states
-
-| Status | Meaning |
-|--------|---------|
-| **Running** | Job is currently executing |
-| **Success** | Completed and result posted to Slack |
-| **Error** | Failed — boss not running, Claude error, or Slack API failure |
+| Service | Description |
+|---------|-------------|
+| **Web** (Next.js 15) | Dashboard — create agents, edit skills, view logs, manage users |
+| **Runner** (Node.js) | Hosts all agent processes and Slack connections |
+| **PostgreSQL** | Stores agents, memories, skills, sessions, users, history |
+| **Redis** | Delivers hot-reload events from Web to Runner instantly |
+
+**How a message flows:**
+1. User @mentions an agent in Slack
+2. Runner receives the event via Bolt Socket Mode
+3. Claude Code processes the message with the agent's compiled `CLAUDE.md`
+4. Agent uses MCP tools if needed (Redshift, GitHub, Notion, etc.)
+5. Response is formatted as Slack Block Kit and posted to the thread
+6. Memory files written during the session are synced to Postgres
+7. Next conversation starts with all accumulated knowledge
 
 ---
 
-## 🔒 Authentication & Roles
-
-SlackHive ships with a simple but effective auth system — no external auth provider needed.
-
-### How it works
-
-- **Superadmin** is configured via environment variables (`ADMIN_USERNAME` / `ADMIN_PASSWORD`) — never stored in the database
-- **Sessions** use HMAC-signed cookies (no JWTs, no session table)
-- **Middleware** protects all routes — unauthenticated requests redirect to `/login`
-
-### Roles
-
-| Role | View | Create/edit agents | Manage jobs | Settings | Manage users |
-|------|------|-------------------|-------------|----------|-------------|
-| **Superadmin** | ✅ | ✅ | ✅ | ✅ | ✅ |
-| **Admin** | ✅ | ✅ | ✅ | ✅ | ✅ |
-| **Editor** | ✅ | ✅ | ✅ | ✅ | ❌ |
-| **Viewer** | ✅ | ❌ | ❌ | ❌ | ❌ |
-
-- **Superadmin**: configured via env vars, never stored in DB
-- **Admin**: full access — can create users with any role
-- **Editor**: can create/edit agents, jobs, MCPs, skills, settings — but cannot manage users
-- **Viewer**: read-only access to everything
-
-All permissions are enforced server-side via API route guards, not just hidden in the UI.
+## 🔑 Claude Code Authentication
 
----
+Two options — use whichever fits your setup:
 
-## 🔌 MCP Server Catalog
-
-MCP servers are managed at the platform level. Add a server once, assign it to any agent.
-
-| Transport | Use Case | Config |
-|-----------|----------|--------|
-| `stdio` | Local subprocess | `command`, `args`, `env` |
-| `sse` | Remote SSE endpoint | `url`, `headers` |
-| `http` | Remote HTTP endpoint | `url`, `headers` |
-
-**Example — Redshift MCP:**
-
-```json
-{
-  "name": "redshift-mcp",
-  "type": "stdio",
-  "description": "Read-only Redshift query access",
-  "config": {
-    "command": "node",
-    "args": ["/path/to/redshift-mcp-server/dist/index.js"],
-    "env": { "DATABASE_URL": "redshift://user:pass@host:5439/db" }
-  }
-}
+**Option A — API Key**
+```env
+ANTHROPIC_API_KEY=sk-ant-api03-...
 ```
+Billed per token via the Anthropic API. Best for teams and production.
 
-Tool names follow the pattern `mcp__{serverName}__{toolName}`.
-
----
-
-## 📁 Project Structure
-
-```
-slackhive/
-├── apps/
-│   ├── web/                        # Next.js 15 — Web UI + API
-│   │   └── src/
-│   │       ├── app/                # Pages, API routes, settings
-│   │       └── lib/
-│   │           ├── db.ts           # Postgres + Redis client
-│   │           ├── auth.ts         # HMAC cookie sessions, bcrypt
-│   │           ├── auth-context.tsx # Client-side auth React context
-│   │           ├── api-guard.ts    # Role guard for API routes
-│   │           ├── boss-registry.ts # Auto-generated boss team registry
-│   │           ├── slack-manifest.ts
-│   │           └── skill-templates.ts
-│   │
-│   └── runner/                     # Agent runner service
-│       └── src/
-│           ├── agent-runner.ts     # Lifecycle manager
-│           ├── claude-handler.ts   # Claude Code SDK integration
-│           ├── slack-handler.ts    # Slack Bolt + Block Kit formatting
-│           ├── compile-claude-md.ts # Skills + memories → CLAUDE.md
-│           ├── memory-watcher.ts   # fs.watch → DB sync (learning)
-│           ├── job-scheduler.ts   # Cron-based scheduled job executor
-│           └── logger.ts           # Structured logging
-│
-├── packages/
-│   └── shared/                     # Shared TypeScript types + DB schema
-│
-├── docker-compose.yml
-├── scripts/dev.sh
-└── .env.example
+**Option B — Claude Pro or Max Subscription**
+```bash
+claude login    # run on host machine, saves credentials to ~/.claude/
 ```
+Mount `~/.claude` into the runner container and leave `ANTHROPIC_API_KEY` unset. Best for individual developers.
 
----
-
-## 🛠 Tech Stack
-
-| Layer | Technology |
-|-------|-----------|
-| **Language** | TypeScript 5.x throughout |
-| **Web UI** | Next.js 15 (App Router), React 19 |
-| **AI Runtime** | Claude Code SDK (`@anthropic-ai/claude-agent-sdk`) |
-| **Slack** | Bolt SDK (Socket Mode) |
-| **Database** | PostgreSQL 16 |
-| **Pub/Sub** | Redis 7 |
-| **Infrastructure** | Docker Compose |
+> Full guide → [slackhive.mintlify.app/configuration/env-vars](https://slackhive.mintlify.app/configuration/env-vars)
 
 ---
 
 ## 🔮 Roadmap
 
-We're actively building and these are on the horizon:
-
-- [ ] **Local model support** — plug in local LLMs via Claude Code's model routing when available
-- [ ] **Agent-to-agent conversations** — agents can directly message each other, not just through Boss
-- [x] **Scheduled tasks** — cron-based agent actions (daily reports, weekly summaries)
-- [ ] **Multi-workspace support** — one platform instance serving multiple Slack workspaces
-- [ ] **Analytics dashboard** — message volume, response times, memory growth per agent
-- [ ] **Webhook triggers** — trigger agent actions from external events (GitHub, Jira, PagerDuty)
-- [ ] **Custom tool builder** — define simple tools in the UI without writing an MCP server
-- [ ] **Agent templates marketplace** — share and import pre-configured agent setups
-- [ ] **Conversation history UI** — browse past conversations and their outcomes in the web UI
-- [ ] **RAG integration** — connect agents to document stores for knowledge retrieval
-
-Have an idea? [Open an issue](https://github.com/amansrivastava17/slackhive/issues) — we'd love to hear it.
+- [x] Boss orchestration + auto-generated team registries
+- [x] Persistent memory system
+- [x] Scheduled jobs
+- [x] Version control with diff view
+- [x] Encrypted environment variables
+- [x] Channel restrictions
+- [ ] Multi-workspace support
+- [ ] Webhook triggers (GitHub, Jira, PagerDuty → agent actions)
+- [ ] Agent-to-agent direct messaging
+- [ ] Analytics dashboard
+- [ ] Custom tool builder (no MCP server needed)
+- [ ] Agent templates marketplace
+- [ ] RAG integration — connect agents to document stores
+
+Have an idea? [Open an issue](https://github.com/pelago-labs/slackhive/issues)
 
 ---
 
 ## 🤝 Contributing
 
-Contributions are very welcome! This project is in active development.
-
 ```bash
-# Clone and install
-git clone https://github.com/amansrivastava17/slackhive.git
-cd slackhive
-npm install
+git clone https://github.com/pelago-labs/slackhive.git
+cd slackhive && npm install
 
 # Start infra
 docker compose up postgres redis -d
 
-# Run services locally
+# Run locally
 cd apps/web && npm run dev      # http://localhost:3000
-cd apps/runner && npm run dev   # Connects to Slack
+cd apps/runner && npm run dev
 ```
 
-Please open an issue before submitting large PRs so we can discuss the approach.
+Open an issue before submitting large PRs so we can align on the approach.
 
 ---
 
-## ⭐ Star History
+## 👥 Contributors
 
-If you find this project useful, please consider giving it a star — it helps others discover it!
+<a href="https://github.com/pelago-labs/slackhive/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=pelago-labs/slackhive" alt="Contributors" />
+</a>
+
+---
+
+## ⭐ Star History
 
-<a href="https://star-history.com/#amansrivastava17/slackhive&Date">
+<a href="https://star-history.com/#pelago-labs/slackhive&Date">
   <picture>
-    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=amansrivastava17/slackhive&type=Date&theme=dark" />
-    <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=amansrivastava17/slackhive&type=Date" />
-    <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=amansrivastava17/slackhive&type=Date" width="600" />
+    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=pelago-labs/slackhive&type=Date&theme=dark" />
+    <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=pelago-labs/slackhive&type=Date" />
+    <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=pelago-labs/slackhive&type=Date" width="600" />
   </picture>
 </a>
 
 ---
 
-## 📄 License
+## 🔒 Security
 
-MIT © 2026 [Aman Srivastava](https://github.com/amansrivastava17)
+Report vulnerabilities to **[aman@pelago.co](mailto:aman@pelago.co)** — please don't open public issues for security bugs. We respond within 48 hours.
 
 ---
 
+## 📄 License
+
+MIT © 2026 [Pelago Labs](https://github.com/pelago-labs)
+
 <div align="center">
-  <sub>Built with Claude Code SDK, Slack Bolt, and a lot of ☕</sub>
+  <sub>Built with Claude Code, Slack Bolt, and a lot of ☕</sub>
 </div>

package/dist/commands/init.js

@@ -15,7 +15,7 @@ const path_1 = require("path");
 const chalk_1 = __importDefault(require("chalk"));
 const ora_1 = __importDefault(require("ora"));
 const prompts_1 = __importDefault(require("prompts"));
-const REPO_URL = 'https://github.com/amansrivastava17/slackhive.git';
+const REPO_URL = 'https://github.com/pelago-labs/slackhive.git';
 /**
  * Runs `slackhive init` — interactive setup wizard.
  *
@@ -62,7 +62,7 @@ async function init(opts) {
     else {
         const spinner = (0, ora_1.default)('  Cloning repository...').start();
         try {
-            (0, child_process_1.execSync)(`git clone ${REPO_URL} "${dir}"`, { stdio: 'ignore' });
+            (0, child_process_1.execSync)(`git clone --depth 1 ${REPO_URL} "${dir}"`, { stdio: 'ignore' });
             spinner.succeed('Repository cloned');
         }
         catch {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "slackhive",
-  "version": "0.1.23",
+  "version": "0.1.25",
   "description": "CLI to install and manage SlackHive — AI agent teams on Slack",
   "bin": {
     "slackhive": "./dist/index.js"