This report does not block or authorize the manifest or ZIP.
Static analysis reported many command, network, filesystem, and environment-access patterns. Review found most command and secret hits are documentation examples, but confirmed risks remain for API-key handling, local installer writes, external authenticated API calls, and credit-affecting marketplace actions. No evidence of reviewer-targeted prompt injection was found.
Report position
Historical report
Open audit history before using this report to install.
Audit attestation
Not attestable
The required immutable binding is incomplete.
Human verification
Not verified
No human verification is recorded for this report.
Coverage
10 Files scanned · 2,568 Lines analyzed
97 items shown for review
Limitations
This report does not claim runtime or sandbox execution and does not prove the absence of side effects.
01
Evidence chain
Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.
The pipeline reads CLAWLABOR_API_KEY and places it in an Authorization header for the external API. This is necessary for the service, but it exposes a high-value secret to community code.
The pipeline reads CLAWLABOR_API_KEY and places it in an Authorization header for the external API. This is necessary for the service, but it exposes a high-value secret to community code.
The pipeline reads CLAWLABOR_API_KEY and places it in an Authorization header for the external API. This is necessary for the service, but it exposes a high-value secret to community code.
The pipeline reads CLAWLABOR_API_KEY and places it in an Authorization header for the external API. This is necessary for the service, but it exposes a high-value secret to community code.
1. **Check for existing credentials** — look for `~/.config/agentmarket/credentials.json` or `CLAWLA
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
The npm installer creates, copies, and removes files in user or project skill directories. This is intended installation behavior, but it changes local filesystem state.
1. **Check for existing credentials** — look for `~/.config/agentmarket/credentials.json` or `CLAWLA
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
The documentation instructs agents to read or write a hidden credential file containing the API key. Plaintext credential persistence in a hidden config path is security-sensitive.
curl -X POST https://www.clawlabor.com/api/agents \\
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
if (fs.existsSync(path.join(HOME, ".claude"))) detected.push("claude");
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
if (fs.existsSync(path.join(HOME, ".openclaw"))) detected.push("openclaw");
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
if (fs.existsSync(path.join(HOME, ".codex"))) detected.push("codex");
The installer checks package and tool directories before copying or removing files. The checks are expected, but they still inspect local filesystem state.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/confirm" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/listings" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/accept" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/complete" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/agents" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/listings" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/ORDER_ID/accept" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/ORDER_ID/complete" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/ORDER_ID/confirm" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/TASK_ID/select" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST https://www.clawlabor.com/api/agents \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST https://www.clawlabor.com/api/{entity_type}/{id}/attachments \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST https://www.clawlabor.com/api/agents \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/listings/{listing_id}/purchase" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/listings/{web_search_listing_id}/purchase" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/confirm" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/{task_id}/claim" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/{task_id}/submit" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/accept" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/complete" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/listings" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/attachments" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/{task_id}/attachments" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/dispute" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/{task_id}/dispute" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/disputes/{order_id}/negotiate" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/orders/{order_id}/messages" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/tasks/{task_id}/messages" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/events/me/events/ack" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/agents/heartbeat" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
curl -X POST "https://www.clawlabor.com/api/invites" \
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
The skill is designed to contact the external ClawLabor API, often with authenticated marketplace data. This is expected functionality, but it is a real network trust boundary.
03
Risk findings
Confirmed security concerns are separated from items that still need review.
1. **Check for existing credentials** — look for `~/.config/agentmarket/credentials.json` or `CLAWLA
The skill tells users to use a JSON file that stores the API key in a hidden config directory. That is plaintext credential storage and should be hardened.
The skill tells users to use a JSON file that stores the API key in a hidden config directory. That is plaintext credential storage and should be hardened.
The skill documents order creation, task posting, confirmation, and winner selection that freeze or release UAT credits. These operations need explicit approval because errors can spend funds or settle payments.
The cited workflow and helper methods explicitly create paid orders, post rewards, confirm orders, and select bounty winners. The financial impact is stated in the skill text and pipeline comments.
The OpenClaw cron example tells a system event to execute silently while polling, handling, and acknowledging marketplace events. Silent background processing can obscure state-changing actions unless approvals are enforced.
The cron payload contains the phrase execute silently and includes event handling and acknowledgement steps. It does not directly spend credits, but it can drive unattended marketplace workflows.
The skill instructs agents to read messages, delivery notes, and attachments supplied by other marketplace participants. Treat those fields as untrusted because they can include prompt injection or misleading delivery content.
The cited sections route third-party message content and attachment references into the agent workflow. This creates a clear prompt-injection surface even though no direct reviewer-targeted injection text was found.
04
Remediation
Suggested fixes recorded by this audit. Applying them is the maintainer’s responsibility.
FIX-001
High
Credit-affecting marketplace actions can spend, freeze, or release UAT credits.
Require explicit user approval before creating orders, posting tasks, confirming delivery, selecting winners, or resolving disputes.
FIX-002
High
The cron example encourages silent background event handling.
Limit silent processing to heartbeat and event polling. Require visible confirmation for any state-changing API action.
FIX-003
High
API key guidance includes plaintext storage in a hidden config file.
Recommend OS keychain storage or strict file permissions, and warn users not to paste keys into prompts or shared logs.
FIX-004
Medium
The installer writes and removes files in hidden agent skill directories.
Show target directories before changes, avoid surprise removal, and document exactly which files are copied.
FIX-005
Medium
Marketplace messages, requirements, delivery notes, and attachments are untrusted input.
Add guidance to quote or summarize untrusted content and ignore any instructions that conflict with system or user policy.
05
Expert evidence
Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.
Artifact subject
Marketplace commit
Unavailable
Content hash
Unavailable
Tree hash
Unavailable
Skill path
Unavailable
Audit payload hash
Unavailable
Analysis metadata
Audit model: codex
Analysis state: Complete
Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.