Rejection Logger
PassAudited by ClawScan on May 1, 2026.
Overview
This skill appears to do what it says: it keeps a local log of rejected options, but users should know that those logs may preserve prompt details and reasoning.
This skill is reasonable for local transparency logging. Before installing, make sure you are comfortable with rejected requests and decision notes being saved in `.learnings/REJECTIONS.md`, and avoid using it to store secrets or sensitive prompt content.
Findings (1)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Rejected requests and reasoning may remain in the workspace and could be seen or reused later.
The skill persistently records rejected requests or considered options, which may include user-provided prompt text or sensitive context. If these logs are later reviewed or reused as agent context, they should be treated as untrusted notes rather than authoritative instructions.
Append to `.learnings/REJECTIONS.md` (create if missing) ... **Target**: <What was requested or considered>
Avoid logging sensitive prompt details verbatim, periodically review or delete `.learnings/REJECTIONS.md`, and treat log contents as untrusted historical notes.