mvpfy-context
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to execute a bundled Node.js script (
analyze-existing-project.mjs) to crawl the local filesystem and generate a project context report.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local project directory.\n - Ingestion points: The script
analyze-existing-project.mjsreads various files (markdown, manifests, docs) from the project root to prepare context.\n - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are included when the agent processes the resulting JSON report.\n
- Capability inventory: The skill utilizes file system access (read/write) and agent-directed shell command execution.\n
- Sanitization: Content excerpts are truncated to 600 characters and simple markdown patterns are removed, but the underlying text content is not sanitized for instructions.\n- [DATA_EXPOSURE]: The skill scans the project directory and reads file contents. Although it attempts to exclude common sensitive directories such as
.gitandnode_modules, it still collects and stores excerpts from project documents and metadata from manifests in a local JSON file that is subsequently read by the agent.
Audit Metadata