From reflection
Applies playbook updates via the playbook-curator sub-agent. Receives reflection output and delegates to curator for counter increments, new bullet creation, and deduplication. Use after running /session-reflect.
How this skill is triggered — by the user, by Claude, or both
Slash command
/reflection:playbook-updateThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
<objective>
<quick_start> Run after /session-reflect to apply playbook changes:
Prerequisites: Run /session-reflect first to generate reflection output. </quick_start>
Load Reflection Output Get the reflection output to process. Accept JSON directly from /session-reflect output Read from file if saved Verify JSON has required fields: session_id, bullets_used, new_bullets Verify bullets_used.helpful and bullets_used.harmful are arrays Verify new_bullets has valid category arrays<step_2>
Delegate to Playbook Curator Use Task tool to delegate playbook updates to curator sub-agent. ``` Task tool parameters: - subagent_type: "playbook-curator" - description: "Update playbook from reflection" - prompt: | Process the following reflection output and update playbook files.Tasks:
1. Increment helpful counters for bullets in helpful list
2. Increment harmful counters for bullets in harmful list
3. Update last_used_session for all used bullets
4. Add new bullets with auto-generated IDs
5. Check for semantic duplicates before adding
6. Return summary of changes made
Full reflection data:
{reflection_json}
</delegation>
</step_2>
<step_3>
<title>Review Curator Output</title>
<description>Review what changes the curator made.</description>
<actions>
<action>If duplicates found, present to user for decision</action>
<action>If conflicts detected, pause for user resolution</action>
</actions>
</step_3>
<step_4>
<title>Commit Playbook Changes</title>
<description>Commit all playbook file changes to git.</description>
<git_commands>
```bash
git add .ai-agents/playbook/*.json
git commit -m "chore: Update playbook from session {session_id}"
</git_commands> </step_4>
<step_5>
Report Summary Provide summary of playbook updates to user. ``` Playbook Updated (Session {session_id})Counters Updated:
New Bullets Added:
Commit: {commit_hash}
</output_format>
</step_5>
</process>
<success_criteria>
<criterion>Reflection output is valid and complete</criterion>
<criterion>Curator sub-agent processes all updates successfully</criterion>
<criterion>Helpful/harmful counters are incremented correctly</criterion>
<criterion>New bullets have unique, sequential IDs</criterion>
<criterion>Duplicate checking is performed</criterion>
<criterion>Changes are committed with descriptive message</criterion>
</success_criteria>
npx claudepluginhub helloworldsungin/ark-ai-agent --plugin reflectionGuides collaborative design exploration before implementation: explores context, asks clarifying questions, proposes approaches, and writes a design doc for user approval.
Creates structured, bite-sized implementation plans from specs or requirements before writing code. Useful for breaking down multi-step tasks into testable steps with file structure and task boundaries.
Synthesizes the current conversation into a structured spec (PRD) and publishes it to the project issue tracker with a ready-for-agent label, without interviewing the user.