ccf-rank
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on executing local Node.js and Python scripts to query the CCF ranking database and rebuild datasets from source files. This is consistent with the skill's stated purpose of providing rank lookups.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources (PDF catalogs) which is then stored in JSON format and queried by the agent. This represents an attack surface where malicious content embedded in the source PDF could theoretically influence the agent's output.
- Ingestion points: Data enters via PDF parsing in scripts/build_ccf_dataset.py and is later read by scripts/query_ccf_rank.mjs.
- Boundary markers: The scripts do not use explicit delimiters or specific instructions to the agent to ignore content within the data fields.
- Capability inventory: The skill includes file read/write capabilities within the project directory and terminal execution of the provided scripts.
- Sanitization: Standard text normalization is performed, but there is no specific filtering for prompt injection patterns in the extracted text.
- [EXTERNAL_DOWNLOADS]: The project includes a dependency on the pypdf library in scripts/build_ccf_dataset.py for PDF processing, which is a standard package available from official registries.
Audit Metadata