Tag: feedback-capture
Structured Task Feedback Capture
Records structured thumbs up/down feedback with context, tags, and rubric scores, featuring automated lesson distillation from chat history to drive prevention rule promotion.
Capture and Apply User Thumbs Feedback
This skill captures explicit user thumbs up or down feedback, requiring a descriptive sentence on success or failure. It also provides mechanisms to run feedback summaries and apply prevention rules before major implementation.
Capture structured agent feedback and rules
This skill captures explicit user quality signals (e.g., thumbs up/down) regarding prior agent output. It structures this feedback into actionable memories and prevention rules, requiring context before promoting the signal to reusable know…
Capturing User Feedback into Structured Memories
Captures explicit user feedback into structured memories and prevention rules to optimise agentic performance. It processes quality signals like thumbs up or down to record repeatable patterns and failure prevention actions.
Structured feedback capture for agentic workflows
This tool allows developers to record structured feedback, including thumbs up/down signals, context, and rubric scores, following task completion or failures. It is designed to distill lessons from chat history and failed tool calls, feedi…
Dual-write skill for capturing user feedback
This skill captures user feedback via a dual-write mechanism, storing data both in Amp's native MCP memory for immediate recall and in ThumbGate for cross-platform analytics and DPO training pairs.
Capture user feedback and apply prevention rules
This skill facilitates the capture of explicit user thumbs feedback and applies defined prevention rules. It is designed to gather qualitative data and validate assumptions before major implementation phases.
Agent Action Feedback and Prevention Gates
ThumbGate captures agent performance feedback (up/down) and uses adaptive mechanisms to generate hard prevention rules. These rules are enforced via PreToolUse hooks, preventing the agent from executing known-bad actions before they occur.