smart-rebase
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple Git commands (git rebase, git merge-base, git diff, etc.) to analyze and manipulate the local repository state. These commands are necessary for the skill's stated purpose.
- [COMMAND_EXECUTION]: The skill instructions allow for the execution of local repository test or lint commands to verify code integrity after conflict resolution.
- [PROMPT_INJECTION]: The skill is subject to an indirect prompt injection surface as it reads and analyzes code and diffs from the repository which could contain adversarial instructions. 1. Ingestion points: File SKILL.md, workflow steps 3, 4, and 7 read data via git diff and git show. 2. Boundary markers: Absent; the agent is instructed to read content without specific markers to isolate untrusted data. 3. Capability inventory: Execution of git commands and repository-defined test/lint scripts. 4. Sanitization: No sanitization is performed on the ingested code content before the agent processes it.
Audit Metadata