verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Feb 20, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- Prompt Injection (LOW): The skill employs strong imperative language and simulated stakes ('The Iron Law', 'If you lie, you'll be replaced') to constrain agent behavior. This is a known prompting technique to reduce hallucinations and ensure task integrity, rather than an attempt to bypass safety or ethical filters.
- Indirect Prompt Injection (LOW): The skill requires the agent to ingest and analyze full output from external commands (tests, linters, builds), which is a vulnerability surface for indirect prompt injection if those tools produce attacker-controlled content. 1. Ingestion points: Command-line output processed during the 'READ' and 'VERIFY' steps. 2. Boundary markers: Absent; no instructions provided to wrap or sanitize external data. 3. Capability inventory: Execution of arbitrary shell commands for verification. 4. Sanitization: Absent; the agent is instructed to read 'Full output'.
- No Code (SAFE): The skill consists entirely of instructional markdown and does not distribute any scripts, binaries, or package manifest files.
Audit Metadata