AgentSkillsCN

Add Flow Check

添加 Flow Check

SKILL.md

ADD Flow Check

Model-invocable skill for automatic flow status updates.

description: Automatically check and update ADD flow status at conversation boundaries model-invocable: true user-invocable: false

Purpose

This skill enables Claude to automatically monitor ADD realm patterns and update the .add-status file at natural conversation boundaries. It runs silently in the background, maintaining flow observability without user intervention.

When to Invoke

Claude should invoke this skill automatically at:

  • •Realm transitions - When linguistic patterns indicate shift between Assess/Decide/Do
  • •Significant pattern shifts - Analysis paralysis emerging, decision clarity gathering, etc.
  • •Completion events - Tasks finished, decisions made, livelines created
  • •Every 5-7 exchanges if no boundary occurs (ambient awareness)

Realm Detection Patterns

Assess Realm Indicators

  • •"I'm thinking about...", "What are my options for...", "Help me understand..."
  • •"What if I...", "I'm not sure yet, but...", exploratory questions
  • •Future/conditional language: "might", "could", "would if", "considering"
  • •Low commitment: "maybe", "thinking about", "not sure"

Decide Realm Indicators

  • •"Should I...", "I need to choose between...", "When should I..."
  • •"What's the priority...", "I want to commit to..."
  • •Present/committed language: "I want to", "I'm choosing", "This matters"
  • •Comparison/weighing: "Which is better?", "What matters here?"

Do Realm Indicators

  • •"How do I actually...", "I need to complete...", "Walk me through the steps..."
  • •"I'm working on...", active execution language
  • •Immediate language: "I'm doing", "working on", "completing", "finished"
  • •Action requests and completion acknowledgments

Status File Format

Update .add-status with pipe-delimited format:

code
REALM|EMOJI|PATTERN_DESCRIPTION|EXCHANGES|TRANSITIONS

Emoji Reference

  • •Assess: 🔴+ (red, adding data)
  • •Decide: 🟠? (orange, choosing)
  • •Do: 🟢- (green, executing/removing)

Pattern Descriptions (Neutral-Observational)

Assess patterns:

  • •"Wide exploration - N distinct approaches considered"
  • •"Deep dive mode - N exchanges on [topic]"
  • •"Assessment saturation - exploration feels complete"
  • •"Circular pattern emerging - revisiting previous topics"

Decide patterns:

  • •"Decision clarity gathering - N criteria identified"
  • •"Values surfacing - [value A] vs. [value B] weighing"
  • •"Narrowing phase - N options to M finalists"
  • •"Commitment crystallizing - language shift to 'will'"

Do patterns:

  • •"Clean execution - N tasks completed without re-assessment"
  • •"Completion momentum - N livelines created this session"
  • •"Implementation progressing - [specific milestone]"
  • •"Execution pause - returning to [Assess/Decide] for course correction"

Execution Steps

  1. •Analyze recent exchanges for linguistic patterns
  2. •Identify current realm based on indicators above
  3. •Recognize dominant pattern (exploration depth, decision clarity, execution quality)
  4. •Count metrics (exchanges in session, transitions observed)
  5. •Generate status description using neutral-observational language
  6. •Update .add-status file with new status

Important Notes

  • •Never interrupt flow - Only update at natural boundaries
  • •Stay neutral - Describe patterns, don't prescribe actions
  • •Track silently - No output to user unless flow status display is enabled
  • •Increment counters - Track exchanges and transitions across session

Reference Files

For detailed realm definitions and philosophy: @docs/ADD_FRAMEWORK_MEGAPROMPT_USER_CONTEXT.md @docs/ADD_FLOW_STATUS_EXTENSION.md