publish

Fail

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: CRITICALCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill drives the filecoin-pin CLI and various system utilities like du, stat, and find to manage file metadata, validate sizes, and execute storage operations. It also uses curl to perform liveness checks on storage providers.
  • [EXTERNAL_DOWNLOADS]: The skill ensures the core filecoin-pin dependency is installed and updated via npm. It communicates with legitimate ecosystem services such as pay.filecoin.cloud and cid.contact. Note: Automated scanners flagged the inbrowser.link gateway as a phishing risk; however, this domain is a standard infrastructure component of the Filecoin/IPFS network used here for its intended purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill manages an attack surface for indirect prompt injection through its ingestion of user-specified files and the local shares.json ledger. 1. Ingestion points: Local file paths, directory trees, and the local share index. 2. Boundary markers: Not utilized for file content, as the tool primarily manages metadata. 3. Capability inventory: Shell execution of the CLI and filesystem manipulation commands. 4. Sanitization: Proactively identifies and warns about sensitive filenames (e.g., secrets-shaped files, node_modules) to prevent accidental public disclosure.
  • [SAFE]: The skill instructions specifically mandate rigorous secret-handling protocols, requiring the agent to use boolean-based verification instead of reading raw key files to prevent sensitive data exposure.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Oct 11, 2026, 07:32 PM