skills/sebas-aikon-intelligence/antigravity-awesome-skills/verification-before-completion/Gen Agent Trust Hub
verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to execute commands (tests, builds, linters) and read their full output to verify status. This process creates a surface where instructions embedded in the output of these commands could potentially influence the agent.
- Ingestion points: Output from verification commands (identified in SKILL.md under 'The Gate Function').
- Boundary markers: No specific delimiters or markers are defined to separate command output from agent instructions.
- Capability inventory: The skill relies on the agent's ability to execute shell commands.
- Sanitization: No sanitization or filtering of command output is instructed before the agent processes it.
- [PROMPT_INJECTION]: The skill uses highly imperative and authoritative language, such as 'The Iron Law,' 'Non-negotiable,' and 'Violating the letter of this rule is violating the spirit of this rule.' These instructions are designed to override default reporting behaviors and enforce strict compliance, mimicking patterns used in prompt injection to control agent output.
Audit Metadata