verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill defines a protocol for executing standard developer commands (e.g., npm test, go build, oxlint) to verify code changes. It requires the agent to capture full stdout, stderr, and exit codes, ensuring that claims like 'tests pass' are supported by verifiable data rather than assumptions.
- [INDIRECT_PROMPT_INJECTION]: The skill involves ingesting data from tool outputs, creating a surface for indirect prompt injection. Ingestion points: Tool output (stdout/stderr) as described in the Gate section of SKILL.md. Boundary markers: The skill defines 'Success Criteria' to guide the agent's interpretation of the output. Capability inventory: The skill relies on shell command execution via agent tools. Sanitization: No explicit output sanitization is mentioned, but the workflow emphasizes evidence-based verification against user claims.
- [SAFE]: No obfuscation (Base64, zero-width chars, homoglyphs), persistence mechanisms, privilege escalation, or data exfiltration patterns were found in the skill metadata, instructions, or examples.
Audit Metadata