git-commit-composer
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the execution of shell commands, specifically
git diff --cachedandgit diff, to retrieve the content of staged changes for analysis. This is a core part of its intended functionality. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data sourced from the local repository's file changes.
- Ingestion points: Data is ingested through the output of the
git diffcommands (referenced in SKILL.md). - Boundary markers: The instructions lack explicit delimiters or instructions to treat the diff content strictly as data, potentially allowing malicious instructions embedded in code comments within the diff to influence the agent's behavior.
- Capability inventory: The agent uses shell commands to read file content. It also generates suggested shell commands for the user to execute (
git reset,git add). - Sanitization: No sanitization, filtering, or escaping is applied to the content retrieved from the git diff before it is analyzed by the language model.
Audit Metadata