claude-reflect

https://github.com/BayramAnnakov/claude-reflect

[{"step":"Gather Feedback","action":"Collect feedback from the relevant source (e.g., emails, Slack threads, PR comments) after completing a task. Ensure the feedback includes specific details about what worked, what didn’t, and any actionable improvements.","tip":"Use Claude's built-in feedback capture by pasting the feedback directly into the chat or referencing a file/URL where it’s stored."},{"step":"Run claude-reflect","action":"Execute the claude-reflect command in your terminal or invoke it via the AI assistant. Specify the task name and feedback source, e.g., `claude-reflect --task 'payment_processor_refactor' --feedback 'client_email.md'`.","tip":"If you're using an AI assistant, you can paste the prompt template and replace the placeholders with your task name and feedback source."},{"step":"Review and Validate","action":"Check the generated CLAUDE.md and AGENTS.md files to ensure the feedback was accurately captured and formatted. Verify that the actionable improvements are clear and specific.","tip":"Look for sections like 'Common Pitfalls' or 'Agent Configuration' in the files to confirm the updates were applied correctly."},{"step":"Take Action","action":"Use the captured feedback to inform your next steps. Prioritize the improvements based on urgency and impact, and share the updates with your team if necessary.","tip":"For team-wide improvements, consider creating a follow-up task or ticket in your project management tool (e.g., Jira, Linear) to track the implementation of the feedback."},{"step":"Iterate","action":"After implementing the improvements, run claude-reflect again to capture any new feedback or lessons learned. This ensures the knowledge base grows continuously.","tip":"Schedule regular reflection sessions (e.g., weekly) to keep the CLAUDE.md and AGENTS.md files up to date."}]