dp-knowledge-compile
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes
subprocess.runin Python andexecFileSyncin Node.js to execute internal utility scripts for project scaffolding, schema generation, and data validation. These commands are used to bridge functionality between Python and Node.js components and target controlled, local script paths. - [EXTERNAL_DOWNLOADS]: The
scripts/fetch_schema.pyscript retrieves ontology schema definitions from a remote Foil instance. The connection details (base URL and credentials) are not hardcoded and must be provided by the user via environment variables. The authentication process includes validation of the remote host to ensure it is a public network address. - [INDIRECT_PROMPT_INJECTION]: The skill ingests XMind files (untrusted data) through
scripts/compile_xmind.pyto produce knowledge artifacts. While it implements sanitization by filtering properties against a declared schema, the resulting data is intended for consumption by AI agents, which is a known attack surface for indirect prompt injection.
Audit Metadata