idea-direction-generator
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local files, creating a surface for indirect prompt injection.
- Ingestion points: The script reads research signals and notes from IDEA_SIGNAL_TABLE.jsonl and paper_notes.jsonl within the workspace.
- Boundary markers: None; the script parses records from structured JSONL files without explicit delimiters for potentially untrusted content.
- Capability inventory: The skill's capabilities are limited to reading from the workspace and writing Markdown and JSONL output to the output/trace/ directory. It lacks network or arbitrary command execution capabilities.
- Sanitization: No explicit sanitization or filtering of the input data is performed within the script before processing.
Audit Metadata