session-feedback
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates entirely on local project data and conversation history to improve agent performance based on user feedback.
- [DATA_EXPOSURE]: No sensitive data access or network exfiltration was detected. The skill writes to standard local directories such as
.claude/memory/or.cursor/skills/for the purpose of maintaining a session memory. - [PROMPT_INJECTION]: The instructions follow a clear workflow for data classification (Corrections, Preferences, Do-differently) and do not attempt to bypass system safety filters or override global agent constraints.
- [COMMAND_EXECUTION]: The skill documentation provides manual installation instructions (copying files) and standard CLI usage patterns. It does not contain any hidden or automated execution of dangerous shell commands.
- [REMOTE_CODE_EXECUTION]: There are no external downloads, piped shell scripts, or dynamic code generation components within the skill.
Audit Metadata