APIClaw — Amazon Commerce Data, 11 Endpoints

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill makes network requests to api.apiclaw.io and apiclaw.io to retrieve commerce data and access documentation. These are the official domains of the skill's author (SerendipityOneInc).
  • [COMMAND_EXECUTION]: The skill includes a Python script (scripts/apiclaw.py) that acts as a CLI for the APIClaw service. It performs network operations and local file I/O for configuration and data aggregation. The documentation also suggests standard environment setup commands (export).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes Amazon product reviews, which are external data sources. The 'Local Review Toolkit' handles this by interpolating review text into prompts for the agent. The implementation includes boundary markers (triple quotes) and strict JSON schema requirements to mitigate the risk of the agent executing instructions found within reviews.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 05:32 PM
Security Audit — agent-trust-hub — APIClaw — Amazon Commerce Data, 11 Endpoints