roadmap-generate
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using Node.js. It specifies running a local script
roadmap-skill/bin/cli.jsor a global commandroadmapsmith. This is legitimate functionality for a CLI-based development tool. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads 'repository evidence' (local project files) to generate or update roadmaps. While this is the intended use case, malicious content placed in the analyzed repository could potentially influence the agent's summary or the generated roadmap content.
- Ingestion points: Processes files within the project directory specified by
--project-root .. - Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded commands within the processed repository files.
- Capability inventory: Executes local CLI tools via shell commands in
SKILL.md. - Sanitization: No mention of content sanitization or validation before processing repository data.
Audit Metadata