integrations-prior-art-survey
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation includes instructions to run a local Python validation script (
scripts/validate_integrations_prior_art.py) usinguv run. This script performs deterministic shape and logic checks on research artifacts and is a standard part of the skill's utility. - [INDIRECT_PROMPT_INJECTION]: The skill identifies a high risk of indirect prompt injection from third-party descriptions in API catalogs and descriptors. It mandates that the agent treat all external content as data only and provides clear instructions to neutralize any instructional prose before recording it.
- [EXTERNAL_DOWNLOADS]: The skill references integration research data from several well-known services and organizations, including GitHub (n8n, Activepieces, Pipedream), npm, PyPI, and Apis.guru. These references are used for data retrieval purposes related to the skill's primary research function.
Audit Metadata