recon-gyms

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard utilities including curl, grep, jq, and sed to perform network reconnaissance, header analysis, and response parsing on user-provided target domains.
  • [EXTERNAL_DOWNLOADS]: Fetches data from external target websites and well-known fitness service providers (e.g., mindbodyonline.com, marianatek.com, clubready.com) to fingerprint platforms and verify API endpoints.
  • [PROMPT_INJECTION]: The skill processes untrusted content retrieved from target websites, creating a potential surface for indirect prompt injection.
  • Ingestion points: Responses from curl commands targeting $TARGET and platform-specific APIs across all steps in SKILL.md.
  • Boundary markers: Absent; the skill does not instruct the agent to use specific delimiters or ignore embedded instructions when interpreting target responses.
  • Capability inventory: Shell execution of curl, grep, jq, and sed to interact with network services and local data.
  • Sanitization: No explicit sanitization or validation of the fetched HTML or JSON content is performed prior to agent processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 05:20 PM
Security Audit — agent-trust-hub — recon-gyms