antithesis-triage
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs the user to install prerequisites from official sources, including the snouty CLI from the vendor's GitHub repository and jq from its official domain. It also fetches test metadata and execution logs from the Antithesis platform using the snouty CLI.
- [COMMAND_EXECUTION]: The skill executes shell commands to interact with the Antithesis API and process data, including snouty (e.g., snouty doctor, snouty runs show, snouty runs logs) and jq for JSON manipulation and log filtering.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze potentially untrusted data from the system under test (SUT) via logs.
- Ingestion points: Metadata and application logs are retrieved via the snouty CLI and processed by the agent in SKILL.md and references/logs.md.
- Boundary markers: The skill does not define specific prompt boundaries for log content, but relies on structured jq filtering to isolate relevant data.
- Capability inventory: The agent has the capability to read local files, execute snouty commands, and run jq for data processing.
- Sanitization: The snouty CLI automatically sanitizes output by stripping ANSI escape codes from log text before it is processed by the agent.
Audit Metadata