From thumbgate
Captures thumbs up/down feedback on agent output into structured memories and prevention rules. Prompts for one-sentence explanation before promoting to reusable memory. Use for explicit quality signals like 'that worked' or 'that failed'.
npx claudepluginhub igorganapolsky/thumbgateThis skill uses the workspace's default tool permissions.
When user provides feedback, execute:
Recalls past mistakes and prevention rules before tasks, captures success/failure feedback after, to prevent agents repeating errors. Uses local MCP tools like recall and capture_feedback.
Detects repeated feedback patterns across Claude sessions and converts them into memories, validation hooks, enforcement patterns, or skills to automate corrections.
Detects corrections and learnings from user messages. Embedded in ask-question skill—do not invoke feedback-capture separately.
Share bugs, ideas, or general feedback.
When user provides feedback, execute:
# negative
node .claude/scripts/feedback/capture-feedback.js \
--feedback=down \
--context="<what failed>" \
--what-went-wrong="<specific failure>" \
--what-to-change="<prevention action>" \
--tags="<domain>,regression"
# positive
node .claude/scripts/feedback/capture-feedback.js \
--feedback=up \
--context="<what succeeded>" \
--what-worked="<repeatable pattern>" \
--tags="<domain>,fix"
If the user only says thumbs up, thumbs down, that worked, or that failed, log the signal and ask one follow-up question before claiming it became reusable memory.
At session start, run:
npm run feedback:summary
npm run feedback:rules