Copied

Open Source · Self-Hosted · 30+ Agents

Your AI team.
One command.

30+ agents pre-installed. Every channel wired. Your team coordinates itself — you just delegate.

MIT Licensed Self-hosted No per-seat fees
Home
Team
Channels
Logs
Good morning, James
Thursday, 9:14 AM
4 agents live
4
Active
30+
Agents
142K
Tokens
🔵
Mike · PR Reviewer ● live
🟢
Scout · Lead Gen ● live
🟣
Counsel · Legal Review ● live

Pre-installed agents — ready the moment you deploy

🔍Researcherresearch
✉️Email Assistantinbox
✍️Doc Writerwriting
💻Coderengineering
📊Analystdata
📱Social Mediamarketing
📅Plannerops
🌍Translatorlanguage
⚖️Legal Assistantlegal
💼Sales Assistantsales
🛡️Security Auditorsecurity
🧑‍💼Recruiterhr
🤝Customer Supportsupport
🧪Test Engineerqa
🏗️Architectsystems
🎓Tutoreducation
🔍Researcherresearch
✉️Email Assistantinbox
✍️Doc Writerwriting
💻Coderengineering
📊Analystdata
📱Social Mediamarketing
📅Plannerops
🌍Translatorlanguage
⚖️Legal Assistantlegal
💼Sales Assistantsales
🛡️Security Auditorsecurity
🧑‍💼Recruiterhr
🤝Customer Supportsupport
🧪Test Engineerqa
🏗️Architectsystems
🎓Tutoreducation
30+ Agents pre-installed
40+ Tools available
5 Channels supported
1 Command to deploy

The Stack

Four services.
Pre-wired. Just run it.

We bundled, configured, and connected OpenClaw, OpenFang, Paperclip, and Hermes. The complex part is already done.

01 — Agent Gateway

OpenClaw

Route messages to the right agent automatically. Discord, Telegram, Slack — all wired from day one.

agent-gateway · channels · routing
Discord connected
Slack connected
Telegram connected
Webhook ready

02 — Agent Runtime

OpenFang

Autonomous agent OS built in Rust. Blazing fast. Each agent runs sandboxed with full tool access — web, code execution, GitHub, and more.

rust · sandboxed · autonomous
Web Search tool
Code Execution tool
GitHub API tool
Browser Agent tool

03 — Orchestration

Paperclip

Workflow orchestration. Cron jobs, agent pipelines, org charts. Model your entire team as versionable YAML code.

orchestration · cron · pipelines
apiVersion: paperclip.ai/v1
kind: AgentPack
agents:
- id: pr-reviewer
trigger: github.pr-opened
model: claude-sonnet

04 — Cognitive Layer

Hermes

Persistent memory, skills, and learning — by Nous Research. Agents that remember context and improve every interaction.

memory · skills · nous-research
Persistent Memory active
Skills Library 48 loaded
Context Window 200K tokens

Why ClawHQ

ClawHQ vs. building it yourself.

The four projects exist. Wiring them together — that's the hard part. We did it.

ClawHQ
DIY setup
30+ agents pre-installed
Included
Find them yourself
All services wired together
Out of the box
Wire manually
Deploy with one command
docker compose up
Days of config
Dashboard with live visibility
Included
Build your own
Cross-service coordination
Native
Custom glue code
Security & audit trails
Built-in
Untracked
Time to first working agent
Minutes
Weeks

Use Case Packs

Pre-wired for every team.

Drop in a pack. Your agents wake up knowing exactly what to do.

Engineering

Autonomous Code Pipeline

Implement features, write tests, review PRs, scan for vulnerabilities, and ship — the full coding loop, end-to-end and unattended.

Code Builder Test Writer Code Reviewer Vuln Scanner Incident Responder

Sales

Pipeline Automation

Lead research, outreach drafts, CRM updates, and deal summaries — zero manual data entry.

Lead Enricher Outreach Writer Deal Tracker

Support

Ticket Intelligence

Auto-triage, draft responses, escalation routing, and weekly trend reports.

Ticket Triage Response Drafter Trend Analyst

Operations

Cross-Functional Coordination

Digest reports from Slack, GitHub, Linear, and email. Surface blockers. Write meeting notes. Keep every team aligned — automatically.

Standup Synthesizer Weekly Digest Blocker Detector Meeting Scribe

Finance

Invoice & Reports

Invoice processing, expense auditing, and board report generation — from 10+ data sources.

Invoice Processor Expense Auditor

HR

Recruiting & Onboarding

Candidate sourcing, onboarding task orchestration, and benefits Q&A — all automated.

Candidate Sourcer Onboarding Bot

Product

Feedback & Intelligence

Feature request triage, NPS synthesis, and competitive monitoring — all from Slack.

Request Triage Competitor Watcher

DevOps Infrastructure

Incident Response & Cloud Cost

Monitor error logs, auto-generate incident checklists, watch cloud spend, and scan GitHub for security vulnerabilities — with Slack approval gates at every step.

Incident Responder Deploy Automator Cloud Cost Optimizer Security Scanner

Enterprise

$40B spent on AI.
95% never reached production.

ClawHQ solves the 8 most critical barriers between AI investment and real business value.

95% of AI pilots never reach production EPAM, 2025
78% struggle to integrate AI with existing systems Zapier, 2025
$40B spent on enterprise AI in 2024 — minimal ROI Deloitte, 2024
$1M+
Data Fragmentation
Avg annual spend on data integration. ClawHQ: multi-system orchestration in weeks.
6→4wk
Time to Production
6-month implementation → 4-week deployment. YAML is versionable, testable, Docker-ready.
70%
User Adoption
70% adoption in 90 days vs. 28% baseline. No-code YAML + Slack interface teams already use.
90d
ROI Timeline
Measurable ROI in 90 days. Pre-built packs generate quick wins. Built-in metrics in Slack.
50%
Audit Burden
Deterministic YAML workflows. Every run logged with inputs, outputs, model version, timestamp.
80%
Fewer Hallucinations
Tool-grounded agents fetch facts from live APIs — no guessing. Multi-model strategy per task.
3–5h
Tool Fragmentation
Hours per person per week saved. One agent handles GitHub + Slack + Discord + email.
70%
Cost vs. SaaS
50–70% savings vs. per-user SaaS at scale. Self-hosted Docker. No per-seat fees. Ever.
Sentinel 6-layer AI security
Prompt injection defense
Canary token compromise detection
SOC 2 fast track
Self-hosted · data never leaves
Full audit trails
Flat per-agent pricing
Human-in-the-loop
Open Paperclip standard
AWS Bedrock, Azure OpenAI & GCP Vertex supported
Use your existing cloud contracts

Enterprise deployment in 4 weeks.

Use your existing AWS Bedrock, Azure OpenAI, or GCP Vertex contract — ClawHQ connects to the providers you already pay for. Your data never leaves your environment.

What teams say

Already changing how teams work.

"Deployed ClawHQ on a Friday. By Monday our PR reviewer had already caught three issues we would have missed. It just works."

👤
Marcus T.
VP Engineering · Series B SaaS

"We replaced 6 different automation tools with one ClawHQ deploy. The consolidation alone was worth it. The AI capabilities were a bonus."

👤
Priya S.
Head of Operations · 200-person team

"Self-hosted was the dealbreaker for our compliance team. ClawHQ was the only option that checked every box — and it's actually good."

👤
Daniel K.
CTO · Regulated FinTech

How it works

Three steps. Then delegate.

01
Deploy

Clone the repo, add your API keys, run one command. Every service starts, connects, and self-configures.

docker compose up -d
02
Meet your team

30+ agents are ready the moment Docker starts. Open the dashboard to see them live — or @mention them directly in Slack.

03
Delegate

Assign tasks, set schedules, or let agents respond to triggers. Your AI team works 24/7 — you review the outputs.

Capabilities

Everything your team needs.

🧠
Multi-Provider LLMs

Claude Haiku, Sonnet, Opus. GPT-4o. Assign the right model to each task for cost and quality.

💾
Persistent Memory

Hermes gives every agent long-term context. They remember prior conversations, company knowledge, and team preferences — across every session and channel.

⚙️
Workflow Automation

Cron schedules, webhook triggers, Slack keyword activation. Agents fire on GitHub events, form submissions, or a message in #engineering. Pipelines chain agents together — output of one becomes input of the next.

📡
Cross-Platform Channels

One agent serves Slack, Discord, Telegram, email, and webhooks simultaneously.

🔒
Sandboxed Execution

Every agent runs isolated. Code execution, browser automation, and API calls — all contained.

📦
Skills Marketplace

Drop in community-built agent packs. Version-controlled YAML — no lock-in.

📊
Cost Dashboard

Real-time token usage, per-agent cost tracking, and budget alerts built in.

Security

Built-in AI agent defense.
Six layers deep.

AI agents face threats that traditional security tools weren't built for. ClawHQ ships with Sentinel — a purpose-built, always-on security system that runs automatically inside every deployment.

01
Gate
Rate limiting, IP reputation, and auth — the perimeter.
02
Prompt Guard
Detects and blocks prompt injection, jailbreaks, system prompt extraction, and indirect injection via tool outputs.
03
Tool Guard
Sandboxes file access, blocks dangerous terminal commands, and catches data exfiltration via tool chains before they execute.
04
Canary Engine
Fake API keys, URLs, and credentials planted in the agent environment. If any are used externally, you have confirmed compromise — not a guess.
05
Decoy Swarm
Honeypot agents in the registry — admin-agent, db-agent, debug-agent. Attackers find them and probe them. You get alerts. They get synthetic noise.
06
Shield
Automated response: block IPs, terminate sessions, enter lockdown, capture forensic snapshots. Alerts to Discord, Slack, email, or PagerDuty.

What Sentinel catches

Prompt injection
Jailbreak attempts
System prompt extraction
Indirect injection via tool outputs
Tool chain exfiltration
Privilege escalation
Malicious shell commands
Reverse shells
Brute force / credential stuffing
MCP server compromise
Canary token exfiltration
Token burn / denial of service
🛡

Sentinel is not a plugin. It is infrastructure — running automatically inside every ClawHQ deployment, zero configuration required.

Read the Sentinel docs →

Privacy

Your data.
Stays yours.

Self-hosted on your infrastructure. Open source. No telemetry. Your agents, your data, your rules.

View source on GitHub →
🏠
Fully self-hosted
Runs entirely on your infrastructure. Nothing leaves your network unless you explicitly send it.
📖
100% open source (MIT)
Every line of code is public. Audit it, fork it, extend it. No black boxes.
🔇
Zero telemetry
No analytics, no usage reporting, no callbacks to our servers. Truly private.
✈️
Air-gap capable
Deploy in fully isolated networks. Compliance, defense, regulated industries — fully supported.

Get started

Your AI team is
waiting for you.

Open source, self-hosted, ready to deploy.