# Clawbridge

> Run Clawbridge discovery from OpenClaw chat

- Skill: `modbender/clawbridge` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add modbender/clawbridge`
- Raw SKILL.md: https://api.skillmd.com/api/skills/modbender/clawbridge/raw
- Safety review: pending (external: skill-scanner WARNING, skillspector CAUTION)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: modbender (https://skillmd.com/u/modbender)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/modbender/clawbridge

---


# Clawbridge Skill

> **Optional chat command** to trigger Clawbridge from OpenClaw.

## What This Skill Does

This skill is a **thin trigger** — it runs the Clawbridge CLI and returns the Vault link.

**The skill does NOT do discovery.** All business logic lives in the runner.

## Behavior

When the user types `/clawbridge`:

1. **Exec**: Run `clawbridge run` locally
2. **Parse stdout**: Extract machine-readable lines:
   - `VAULT_URL=...`
   - `CANDIDATES_COUNT=...`
   - `DISCOVERY_SOURCE=...` (optional)
3. **Reply in chat**:
   - "Done — found X candidates."
   - "Review here: <vault url>"

## Usage

```
/clawbridge
```

Or with a profile:

```
/clawbridge --profile myprofile
```

## Example Output

```
Done — found 3 candidates.

Review here: https://clawbridge.cloud/app/workspaces/xxx/runs/xxx
```

## Prerequisites

**Don't have Clawbridge yet?** Get started at:

👉 **https://clawbridge.cloud**

1. Create an account
2. Create a workspace
3. Follow the setup instructions

Or if you already have an account:

```bash
# 1. Install runner
curl -fsSL https://clawbridge.cloud/install | bash

# 2. Link workspace (get code from your workspace page)
clawbridge link CB-XXXXXX
```

## Architecture

```
User: /clawbridge
        │
        ▼
┌───────────────────────────────┐
│  Skill: exec clawbridge run   │
└───────────────────────────────┘
        │
        ▼
┌───────────────────────────────┐
│  Runner: Discovery workflow   │
│  - Build prompts (private)    │
│  - Call OpenClaw as worker    │
│  - Upload to Vault            │
│  - Print VAULT_URL=...        │
└───────────────────────────────┘
        │
        ▼
┌───────────────────────────────┐
│  Skill: Parse + reply         │
└───────────────────────────────┘
```

## Mental Model

- **Runner = product** (owns discovery strategy, prompts, ranking)
- **Web = vault + approval** (review candidates, approve outreach)
- **Skill = chat shortcut** (optional convenience)

You don't need this skill to use Clawbridge. Run `clawbridge run` directly from terminal.