pathfinder
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted codebase content, including source code, READMEs, and project documentation, to synthesize architectural flowcharts and unified system designs.
- Ingestion points: Phase 0 and Phase 1 involve subagents walking the source tree and reading file content verbatim.
- Boundary markers: The instructions lack explicit boundary markers or warnings to ignore embedded instructions within the content read from the codebase.
- Capability inventory: The agent is authorized to perform file discovery, file reading, and writing artifacts (markdown files and directories) to the repository root.
- Sanitization: No sanitization process is described for handling potentially malicious comments or strings within the analyzed code that could influence the synthesis process.
- [COMMAND_EXECUTION]: The skill directs the agent to identify and document side effects such as process spawns, HTTP calls, and database writes. Additionally, the orchestrator writes a structured set of artifacts to the repository root in a directory named
PATHFINDER-<YYYY-MM-DD>.
Audit Metadata