The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/run.py uses dynamic path calculation to locate the repository root and modifies sys.path to import local utility modules. This enables the skill to access shared project libraries but involves runtime modification of the execution environment.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests data from external JSONL files and incorporates that data into markdown reports without sanitization. 1. Ingestion points: scripts/run.py reads direction data from IDEA_DIRECTION_POOL.jsonl and IDEA_SCREENING_TABLE.jsonl. 2. Boundary markers: The skill does not use specific delimiters or instructions to prevent the agent from following directions potentially embedded within the input data. 3. Capability inventory: The script performs file system operations (write_markdown, write_jsonl) in scripts/run.py to save the results to the workspace. 4. Sanitization: No escaping or validation is performed on the data fields before they are interpolated into the rationale templates.

idea-shortlist-curator