verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill consists of markdown instructions that guide the agent to perform verification steps using standard development tools like npm, pytest, and go. These instructions are designed to prevent premature completion claims and ensure reliability.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to execute commands and "Read the full output" to verify success. This creates a standard ingestion surface for data that could potentially contain indirect prompt injections if the command output (e.g., test logs or build logs) is influenced by an attacker. This is an inherent risk of the verification task and is documented here for completeness.
- Ingestion points: Command outputs from shell execution (e.g.,
npm test,pytest,go test,npm run build) and network responses (e.g.,curl -I) as described inSKILL.md. - Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded commands within the output being verified.
- Capability inventory: The agent is expected to run shell commands and perform network operations.
- Sanitization: No sanitization or validation of the command output is mentioned before the agent processes it.
Audit Metadata