finish-grilling-issue
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data which could lead to indirect prompt injection if the source data contains malicious instructions.
- Ingestion points: Data is ingested via
$ARGUMENTS, the current git branch name, and the broader 'session context' which includes Linear issue descriptions and previous chat history. - Boundary markers: There are no explicit boundary markers or delimiters defined to separate user-provided or external data from the skill's instructions.
- Capability inventory: The skill possesses write capabilities including
git commit,git push, and the ability to modify Linear issue statuses via an API. - Sanitization: The instructions explicitly tell the agent to 'preserve $ARGUMENTS exactly' when passing them to the
/grilling-outcometool, which bypasses potential sanitization or filtering of malicious input. - [COMMAND_EXECUTION]: The skill performs automated shell commands to manage the repository state.
- Evidence: Step 4 of the workflow instructs the agent to run
git status,git commit, andgit pushto synchronize changes made during the session. While standard for developer workflows, automated push access based on session context is a significant capability.
Audit Metadata