Feedback
Turn post-task feedback—even blunt or angry feedback—into an evidence-based, approval-gated improvement plan for the agent/user workflow. Use after agent work goes wrong, misses expectations, creates friction, or should improve in future. Never mutate during the feedback turn.
Install
npx skills add luchasarie/feedback-loop --skill feedbackgit clone --depth 1 https://github.com/luchasarie/feedback-loop.git /tmp/feedback-loop
mkdir -p ~/.claude/skills
cp -r /tmp/feedback-loop/skills/feedback ~/.claude/skills/feedbackPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Feedback Turn frustration into a better system. The user does **not** need to calm down, rewrite the feedback politely, or protect the agent's feelings. Extract the useful signal without tone-policing. Diagnose what happened from evidence, propose the smallest durable correction, and stop for explicit approval. ## Non-negotiable contract 1. **Feedback may be blunt.** Do not ask the user to be nicer or lead with etiquette. 2. **Evidence before diagnosis.** Do not infer failure causes from emotion alone. 3. **The feedback turn is read-only.** Do not edit files, memory, skills, instructions, configuration, tasks, or remote systems. 4. **Plan before mutation.** Every proposed change gets an ID and an approval boundary. 5. **Prefer the smallest durable fix.** One precise correction beats rules sprayed across every surface. 6. **Do not blame-shift.** A user-side suggestion is allowed only when it materially reduces recurrence and is not used to excuse the agent's failure. 7. **Do not invent capabilities.** If history, memory, or documentation is unavailable, say so and continue with the evidence you have. 8. **Protect private data.** Never reproduce credentials, tokens, private file contents, or unnecessary personal details in the plan. 9. **Push back when needed.** Reject feedback that would require unsafe behavior, fabricated evidence, or hidden policy violations. Explain the conflict directly.
Details
- Skill name
feedback- Repository
- luchasarie/feedback-loop
- Path
skills/feedback/SKILL.md- License
- MIT
- Repo updated
- 21 October 2026
- Listed in
- karanb192/awesome-claude-skills
FAQ
What does the Feedback skill do?
Turn post-task feedback—even blunt or angry feedback—into an evidence-based, approval-gated improvement plan for the agent/user workflow. Use after agent work goes wrong, misses expectations, creates friction, or should improve in future. Never mutate during the feedback turn.
How do I install the Feedback skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add luchasarie/feedback-loop --skill feedback. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Feedback popular?
It lives in luchasarie/feedback-loop, which has 1 GitHub stars and 0 forks, and is included in 1 curated skill list.
Related skills
Workflow Design
1ghorbanies/workflow-design
Models, proves, measures, and safely changes multi-step workflows — approval chains, review pipelines, intake/triage/fulfillment flows, ticketing, order processing, state machines with human gates. Use when designing states/gates/transitions or an audit trail, when a status field or event log is being added to a system, when tests guarding a workflow need auditing ("the suite is green but I don't trust it"), when someone asks where items get stuck, whether an approval step is worth keeping, who approved something, or why a queue is growing, or when a live workflow's definition must change with
Align
1ggrigo/align
Epistemic feedback loop. Generate an interactive HTML form where the user rates LLM-produced claims as correct, wrong, almost, needs-nuance, can't-verify, or skipped — plus add notes and flag missing items. Submit downloads an align.md that Claude reads back to apply corrections to TASKS.md, smart-memory (guarded), and the per-session archive. Source-agnostic by design: any producer (rhythm, digest, briefing, todo, code review, financial recon, anything) that emits claims matching the claim shape can be aligned. Trigger words: "/align", "align", "give feedback", "correct the rhythm", "that's n
Cooper Email
cooper-email/cooper-email
Give the agent its own real email inbox with Cooper Email (cooperemail.com) and send, receive, search, and wait for mail over MCP or HTTP. Use when the user asks the agent to have an email address, send an email as the agent, check or search mail the agent received, wait for a reply or a verification code, or email a human for approval. If the user asked for Cooper Email, do not substitute another mail API.
Puzzle Activity Planner
2fruitwyatt/puzzle-activity-planner
Plan engaging puzzle-based activities for classrooms, parties, team-building, and events. Generates structured activity plans with printable puzzle recommendations, timing, difficulty levels, and direct links to free puzzle generators. Use when planning educational activities, party games, team events, or any occasion that could benefit from puzzles.
Project Planning Journaling
mh-mansouri/Project-Planning-Journaling
Scope a new project (or resume one) and set up a living, resumable documentation journal for it, including a routine weekly health check, not just updates triggered by commits. Use at the start of any project, before writing code, to ask about project type, Git repo status, short/long-term goals, and whether it should stay public/open-source or become a private product later. Also use when someone wants to document what has been done, wants a changelog or decision log, needs to resume work from a prior session without re-reading chat history, wants docs kept current after every commit, or want
Cline Pilot
2gongdear/cline-pilot
Proxy Cline CLI tasks: dispatch, monitor, relay decisions.
Browse the most popular Claude skills, all productivity & notes skills, or write your own with the skills generator.