sdd-explore

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of exploration topics which guide its codebase investigation, creating a potential surface for instructions hidden within the target codebase or topic description to influence the agent.
  • Ingestion points: The skill receives a 'topic or feature to explore' from the orchestrator as documented in SKILL.md.
  • Boundary markers: There are no explicit delimiters or boundary instructions defined to isolate the user-provided topic from the agent's internal investigative instructions.
  • Capability inventory: The skill has the capability to read entry points, search for functionality, and check existing tests across the codebase. It also writes to exploration.md.
  • Sanitization: The skill does not perform sanitization or validation of the input topic before using it to perform file system searches and reads.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 10:38 PM
Security Audit — agent-trust-hub — sdd-explore