project-discovery
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands such as
grep,find,ls, andcatto analyze the file structure and content of a target project repository. These commands are used to extract configuration, dependencies, and business logic for the purpose of generating documentation. - [INDIRECT_PROMPT_INJECTION]: The skill reads and interprets content from untrusted external sources, specifically the target project's repository (e.g., source code, README files, and existing documentation). This creates an attack surface where malicious instructions embedded in the target project could influence the behavior of the AI agent during the discovery process.
- Ingestion points: Target project repository files,
README.md, existing.context/files, and user-supplied business or domain documentation. - Boundary markers: The skill does not implement explicit boundary markers or delimiters (e.g., XML tags or "ignore instructions" warnings) when interpolating external content into its prompts.
- Capability inventory: The skill executes shell commands (
ls,cat,grep,find,jq) and writes to the local filesystem (under the.context/directory). - Sanitization: The skill lacks explicit sanitization or filtering logic for the data ingested from the target project before it is processed by the AI.
Audit Metadata