ticket-fix-flow
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is subject to an indirect prompt injection surface (Category 8) as it ingests data from external observability platforms and user descriptions to generate tickets. Ingestion points include Step 1 of both Observability and Free-form modes. While explicit boundary markers are not used when interpolating this data into descriptions, the skill implements sanitization for titles and its capabilities are limited to ticketing and git branch management, presenting minimal risk.
- [COMMAND_EXECUTION]: The skill executes local git commands (git log, git checkout) to gather project context and create feature or fix branches. These operations are restricted to standard developer workflows.
- [EXTERNAL_DOWNLOADS]: The skill interacts with established third-party observability and ticketing platforms (such as Sentry, Honeybadger, GitHub, and Jira). No unauthorized remote code execution or suspicious external downloads were identified.
Audit Metadata