plan-docs-sync
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill methodology in
references/drift-taxonomy.mdsuggests running local project scripts vianpm run <cli> -- --help. Executing scripts from an untrusted repository could lead to malicious code execution if the project's package scripts are compromised. - [INDIRECT_PROMPT_INJECTION]: The skill processes extensive project data, including documentation and source code, creating a surface for indirect prompt injection.
- Ingestion points: Files matching
README.md,CONTRIBUTING.md,docs/**/*.md, and source files (*.ts,*.py) are ingested for auditing as specified inSKILL.mdandreferences/drift-taxonomy.md. - Boundary markers: The instructions lack explicit semantic delimiters or instructions for the model to ignore embedded commands within the analyzed documentation content.
- Capability inventory: The skill possesses capabilities for file system reads and shell command execution (
npm run,grep,glob) as documented in the taxonomies. - Sanitization: No specific sanitization or escaping of external project content is described before the data is processed by the agent.
Audit Metadata