codebase-context

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill utilizes a restricted set of read-only shell commands (ls, cat, find, grep, wc) to inspect project organization and metadata, which is appropriate for its stated objective of codebase analysis.\n- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests content from external files within the analyzed codebase.\n
  • Ingestion points: External project files like package.json, requirements.txt, and README.md are read using cat and head commands.\n
  • Boundary markers: No explicit delimiters or instructions to ignore instructions within the read content are present in the skill definition.\n
  • Capability inventory: The skill possesses the ability to execute find, head, cat, ls, and wc via Bash, and uses the Grep and Read tools.\n
  • Sanitization: No sanitization or validation of the content of the read files is performed before the data is passed to the agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 06:52 AM