recon-scope-triage
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data sources like ASM reports and breach combos, which could contain malicious instructions designed to manipulate agent behavior during triage.\n
- Ingestion points: Recon exports, breach corpora, and mobile app metadata in SKILL.md.\n
- Boundary markers: Includes a mandatory confirmation gate requiring explicit user verification of targets and authorization.\n
- Capability inventory: Utilizes curl for network requests, cat for reading files, and cmp for data comparison.\n
- Sanitization: Recommends verifying content signatures and using control paths to detect catch-all server responses.\n- [COMMAND_EXECUTION]: The skill provides bash templates using curl and cmp for active network interaction and file analysis. These commands constitute an active probing capability that relies on user-supplied target parameters.
Audit Metadata