The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to retrieve and display data from scout runs, including session logs, reports, and agent memory. This data may contain content from external project sources that could harbor malicious instructions intended to influence the exploring agent.\n- Ingestion points: Data is ingested via tasks-runs-session-logs-retrieve, inbox-reports-retrieve, and scout-scratchpad-search.\n- Boundary markers: The skill instructions and scripts like scripts/render_run_report.py help structure the output by distinguishing between agent narration and tool outputs, but do not provide explicit instructions to ignore natural language commands within the processed data.\n- Capability inventory: The agent has access to PostHog project management and monitoring tools, and can execute local Python scripts for data formatting.\n- Sanitization: The provided Python utility scripts perform formatting and text wrapping but do not implement sanitization or filtering for natural language instructions embedded in the data.

exploring-scouts