declare

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from codebase files.
  • Ingestion points: Step 1 and Step 2 instructions direct the agent to read README, package.json, pyproject.toml, go.mod, and other configuration files from the project codebase.
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands are provided for the scanned content.
  • Capability inventory: The skill allows the agent to write an interf.yaml file (SKILL.md) to the project root and provide a natural language summary.
  • Sanitization: No sanitization or validation of the extracted data is specified before it is used to generate the contract.
  • [NO_CODE]: The skill consists entirely of markdown instructions and does not include or reference any executable scripts, binaries, or external code packages.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 10:03 PM