assimilate-popular-workflows
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from external sources, specifically SKILL.md files and READMEs from public GitHub repositories. This creates a surface for indirect prompt injection where malicious instructions in those files could influence the agent's behavior during analysis.\n
- Ingestion points: Public GitHub repositories and the ClawHub marketplace.\n
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when reading untrusted content.\n
- Capability inventory: Uses the GitHub CLI for authenticated searches, clones repositories, and writes generated research to the local filesystem.\n
- Sanitization: No explicit sanitization of external content is mentioned.\n- [COMMAND_EXECUTION]: Utilizes the GitHub CLI (gh) for searching and API interaction. These commands run within the user's environment to fetch data from the internet.\n- [EXTERNAL_DOWNLOADS]: Fetches data from vendor-associated resources on ClawHub and performs clones of public repositories from GitHub for research purposes.\n- [DYNAMIC_EXECUTION]: The skill generates JavaScript (.cjs) process files by codifying logic extracted from untrusted external sources. While the workflow includes a manual review stage before these processes are promoted to the library, the generation phase involves translating external logic into executable templates.
Audit Metadata