researching

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXPOSURE]: The skill reads codebase files to generate technical documentation. While this is its core function, it constitutes an ingestion point for potentially untrusted data from the repository being analyzed.
  • [COMMAND_EXECUTION]: The skill spawns parallel sub-agents (codebase-locator, codebase-analyzer, etc.) to perform research tasks. These sub-agents are given specific scopes but operate autonomously based on the synthesized plan.
  • [PROMPT_INJECTION]: The skill is subject to indirect prompt injection (Category 8) via the files it reads.
  • Ingestion points: Codebase files mentioned by the user or discovered during research are read in their entirety (SKILL.md, Step 1).
  • Boundary markers: No specific delimiters or safety instructions are defined to prevent the agent from following instructions embedded in code comments or documentation.
  • Capability inventory: The agent can read files, write to the filesystem (thoughts/ directory), and spawn additional task-based agents.
  • Sanitization: There is no evidence of sanitization or filtering of the content read from files before it is processed or used to generate new research documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 10:51 PM