maintain-verification-skill
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes project source code and feature documentation to identify drift and generate verification recipes. Maliciously crafted content within the project source (e.g., in comments or README files) could potentially influence the agent's analysis or the recipes it creates.\n
- Ingestion points: Project source files,
SKILL.md, and thefeatures/directory.\n - Boundary markers: The instructions lack explicit delimiters or instructions to ignore embedded commands in the source data.\n
- Capability inventory: The skill spawns concurrent subagents for source analysis, executes local harness scripts, modifies files in the skill directory, and creates Pull Requests.\n
- Sanitization: No sanitization or validation of external content is described.\n- [COMMAND_EXECUTION]: The skill executes local "harness scripts" to drive live application states for feature verification. This involves running project-local executables to interact with UIs, CLIs, or servers.\n
- Evidence: The skill instructions in Step 4 and Step 5 explicitly mention executing harness scripts and following the launch model to drive application features.
Audit Metadata