get-transit
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation instructs the user to install the 'astro-ascendant' package from PyPI if it is not already present. This is a legitimate dependency provided by the vendor for the skill's functionality.
- [COMMAND_EXECUTION]: The skill operates by executing a bundled Python script ('scripts/get-transit.py') with arguments for name, date, and chart division derived from user input.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing and displaying content from external data files.
- Ingestion points: The script reads data from 'persons//CONTEXT.md' and 'persons//charts/D1.json' using the 'PersonRecordStore' class.
- Boundary markers: Although the script includes source citations (e.g., '[sources: ...]') in the generated report, it does not use strong delimiters or explicit instructions to prevent the agent from obeying commands embedded within the data.
- Capability inventory: The skill executes its own script and generates a report intended for downstream processing by the 'parashari-judgement' evaluator, forming a multi-step instruction chain.
- Sanitization: There is no evidence of content sanitization or validation for the data ingested from the external markdown and JSON files before it is rendered in the final output.
Audit Metadata