verification-before-completion

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes a workflow where the agent must read and analyze the full output of arbitrary commands (such as tests, linters, or builds). This creates a surface for indirect prompt injection if the output of these external commands is manipulated by an attacker to include malicious instructions.
  • Ingestion points: The agent is instructed to "READ: Full output" of commands in 'The Gate Function'.
  • Boundary markers: No specific delimiters or instructions to ignore embedded prompts within the command output are defined.
  • Capability inventory: The skill assumes the ability to execute project-relevant commands (tests, builds, linters) and read their results.
  • Sanitization: The skill does not provide methods for escaping or validating the content of the command output before processing.
  • [NO_CODE]: The skill file contains metadata and natural language instructions for the agent but does not include any accompanying scripts, binaries, or executable code blocks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 06:55 AM
Security Audit — agent-trust-hub — verification-before-completion