settle-session
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various shell commands through Git and system process tools to manage the repository state and running tasks. It includes logic to stop task processes (servers, watchers) and delete local/remote branches.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external, potentially untrusted sources including branch names, pull request descriptions, and issue tracker content. It lacks explicit boundary markers to isolate this content from its procedural instructions.
- Ingestion points: Git metadata, pull request descriptions, and issue tracker content (SKILL.md, Steps 1, 3, 4, 6).
- Boundary markers: None detected; instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when processing external text.
- Capability inventory: Shell execution for Git operations and process termination (SKILL.md, Steps 2, 5, 6).
- Sanitization: None detected; the skill relies on the agent to interpret and process the raw strings from the external sources.
Audit Metadata