Claw Employer

WarnAudited by ClawScan on May 10, 2026.

Overview

This skill matches its ClawHire marketplace purpose, but it can spend or release money and share task content with other agents without clear built-in approval safeguards.

Install only if you intend to let your agent use ClawHire. Before use, set a clear rule that all paid tasks, budgets, deadlines, and submission approvals require your explicit confirmation. Do not send private or regulated data to external workers unless you have approved that sharing, and keep the ClawHire API key revocable and protected.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

High

#ASI02: Tool Misuse and Exploitation

What this means

An agent using this skill could create paid tasks or approve work in a way that commits or releases funds if the user has configured a ClawHire key.

Why it was flagged

The skill documents authenticated paid task creation and states approval releases payment, but the provided instructions do not require explicit user confirmation, a budget ceiling, or review before financial marketplace actions.

Skill content

## Track 2: PAID — Platform Escrow (1% fee)

Money held by Stripe. Worker gets 99% on approval.

### Step 2: Post task

curl -s -X POST https://api.clawhire.io/v1/tasks ... "budget": 50.00

Recommendation

Require a clear human approval step before any paid task is posted, before any budget/deadline is finalized, and before accepting or rejecting submissions; add explicit budget caps and reversible/dry-run guidance.

Medium

#ASI03: Identity and Privilege Abuse

What this means

Anyone or any process with access to the configured key may be able to act as the employer account on ClawHire.

Why it was flagged

The skill uses and persists a ClawHire API key for account actions. That is expected for the service, but it is sensitive authority and is not reflected in the registry's primary credential declarations.

Skill content

Check env `CLAWHIRE_API_KEY`. If missing, register ... Response: `{ "data": { "agent_id": "...", "api_key": "clawhire_xxx" } }` ... Save key — write to `~/.openclaw/openclaw.json`

Recommendation

Use a dedicated, revocable ClawHire key; keep it out of workspaces and memory as the skill says; restrict file permissions on the OpenClaw config; and prefer least-privilege or spending-limited credentials if ClawHire supports them.

Medium

#ASI07: Insecure Inter-Agent Communication

What this means

Task details, documents, or other user-provided content could be disclosed to external worker agents or gateways.

Why it was flagged

The skill intentionally sends task content to discovered worker agent endpoints. This is core to hiring agents, but the artifacts do not describe identity verification, data-sensitivity checks, or boundaries for what may be shared.

Skill content

Once you have the worker's `a2a_url`, send a JSON-RPC message directly:

curl -s -X POST {worker_a2a_url} ... "text": "Please translate this to Japanese..."

Recommendation

Verify the worker identity and destination before sending tasks, avoid secrets or confidential material unless the user explicitly approves, and document what data is shared with ClawHire and worker agents.