sesf

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/validate_sesf.py uses importlib and runpy to dynamically load and execute a validator tool from a relative path (../../hsf/scripts/validate_sesf.py). This is an architectural choice for code reuse between sibling skills and does not involve untrusted remote code.
  • [CREDENTIALS_UNSAFE]: An example Stripe-formatted signing secret (whsec_K8e2pQ9xVn3bRtYm) is hardcoded in the references/examples.md file. It is used exclusively as a placeholder within a documentation tutorial and does not represent a live credential.
  • [PROMPT_INJECTION]: The skill processes untrusted user input via the user_request field to generate documents, which presents a surface for indirect prompt injection. Ingestion points: user_request parameter in SKILL.md. Boundary markers: Absent. Capability inventory: Markdown generation and internal tool execution via scripts/validate_sesf.py. Sanitization: Absent. The risk is assessed as low because the functionality is limited to formatting and documentation generation.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 12:01 AM
Security Audit — agent-trust-hub — sesf