skills/q00/openclip/oc-cut-proposer/Gen Agent Trust Hub

oc-cut-proposer

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface where it ingests untrusted external data that could potentially contain malicious instructions targeting the agent's behavior.
  • Ingestion points: The skill reads from <PROJECT>/transcript.json and <PROJECT>/analysis.json as specified in the SKILL.md instructions.
  • Boundary markers: The instructions do not define specific delimiters or "ignore instructions" warnings to separate user-provided transcript text from the agent's internal logic.
  • Capability inventory: The skill is configured with access to the Bash and Read tools, providing file system access and shell command execution capabilities.
  • Sanitization: There are no documented steps for sanitizing, validating, or filtering the content read from the external JSON files before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 10:33 PM
Security Audit — agent-trust-hub — oc-cut-proposer