Agent skill
user-feedback
Use when asked to check, read, triage, or act on user feedback
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/user-feedback
SKILL.md
User Feedback Protocol
Use this skill to keep USER_FEEDBACK.md in sync with actionable tasks. Monitor and process USER_FEEDBACK.md in the workspace root: check if new feedback exists, turn actionable feedback into tasks using the repo's preferred task system, and update the read timestamp after processing.
File format
- Line 1: last-read timestamp in local time, format
YYYY-MM-DDTHH:MM:SS. - Line 2: human note: Do not delete the timestamp above; it records the last time this file was read by an agent.
- Lines 3+: free-form user feedback.
Commands
Use the bundled script:
python3 ~/.codex/skills/user-feedback/scripts/user_feedback.py check_user_feedback
python3 ~/.codex/skills/user-feedback/scripts/user_feedback.py init_user_feedback
python3 ~/.codex/skills/user-feedback/scripts/user_feedback.py update_user_feedback
Check
- Run
check_user_feedbackfrom the workspace root (defaults to./USER_FEEDBACK.md). - If it exits with code 2 and stderr contains
File not found, runinit_user_feedback. - If it returns
false(exit code 1), ignore the file and stop. - If it returns
true(exit code 0), open the file and process feedback.
Process feedback
- Read lines 3+ for actionable items (fixes, requests, bugs, improvements, questions).
- Create one task per actionable item using the repo's preferred task system.
- If a
.tasks/directory exists, use thesvtask workflow. - If a
.tickets/directory exists, use thetkticket workflow. - If neither exists, look for repo instructions (AGENTS.md) or ask the user.
- If a
- Keep each task concise and specific; include any needed context from the feedback.
- Check existing tasks for overlap and duplicates, especially if they are still open.
Update
After all actionable feedback is represented in tasks, run update_user_feedback to update the timestamp and mark the file as read.
Script details
The script lives at scripts/user_feedback.py inside this skill and implements:
check_user_feedback(compares file mtime vs first-line timestamp with a 1s threshold)init_user_feedback(creates the file with timestamp + note)update_user_feedback(replaces first line with current timestamp)
Use --path to override the default file path and --force to overwrite on init.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?