rerun-blueprint
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external recording data and schemas which could potentially contain instructions designed to influence agent behavior.\n
- Ingestion points: The skill reads metadata and schemas from Rerun recording files (.rrd) and datasets using RrdReader, rerun_get_recording_schema, and rerun_get_dataset_schema as described in SKILL.md.\n
- Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the processed recording data.\n
- Capability inventory: The skill instructions include process spawning (ViewerClient.spawn), network connectivity (rec.connect_grpc), file writing (viewer.save_screenshot), and system state modification (rerun_set_blueprint).\n
- Sanitization: No sanitization or validation logic is specified for data retrieved from external recordings before it is processed or displayed.\n- [COMMAND_EXECUTION]: The skill provides instructions and code examples for executing local processes and scripts.\n
- The SKILL.md file contains code utilizing ViewerClient.spawn to start a local Rerun viewer process.\n
- The frontmatter configuration allows the use of the Bash tool for running development and visualization scripts.
Audit Metadata