Delegate independent tasks to specialized agents working concurrently with isolated context.
Meta-Chain-of-Skills 150 Protocol
01-meta-chain-of-skills-150
[01] META. Сканирует доступные skills, создает план выполнения и идет шаг за шагом с подтверждением каждого этапа. Triggers on complex tasks, multi-step work, or when structured execution is needed.
SKILL.md
Full skill instructions
Meta-Chain-of-Skills 150 Protocol
Core Principle: Найди skills → создай план → выполняй по шагам → подтверждай каждый шаг.
What This Skill Does
- SCAN: Сканирует систему на доступные skills
- BUILD: Создает план из подходящих skills
- EXECUTE: Выполняет по одному этапу с подтверждением
- REROUTE: Может перестроить план если нужно
Three Steps
Step 1: SCAN (Find Available Skills)
🔍 SCANNING SYSTEM...
Looking for skills in:
- ./.codex/skills/ (project skills, canonical)
- ./skills/ (legacy project skills)
- ./.claude/skills/ (legacy project-specific)
- ~/.claude/skills/ (personal skills)
Found skills: N total
Available for chaining:
- [skill-1] — [purpose]
- [skill-2] — [purpose]
- [skill-3] — [purpose]
...
Step 2: BUILD (Create Chain)
📋 BUILDING PLAN...
Query: [user query]
Goal: [identified goal]
Plan created:
1. [skill-name] — [purpose]
2. [skill-name] — [purpose]
3. [skill-name] — [purpose]
...
Total steps: N
✅ Start execution? (Yes / Modify / Cancel)
Step 3: EXECUTE (Run Stages)
Each stage follows this format:
⚡ EXECUTING — Step N of X
Current skill: [skill-name]
Purpose: [what this stage achieves]
[... skill execution output ...]
───────────────────────────────────────────────────────────────
📊 EXECUTION STATUS
✅ Completed: Step 1, Step 2, ...
⚡ Current: Step N — [skill-name]
⏳ Remaining: Step N+1, Step N+2, ...
📍 Progress: ████████░░░░░░░░ N/X steps (XX%)
🎯 Next: Step N+1 — [next-skill-name]
Purpose: [what next stage will do]
Continue to next step? (Yes / Pause / Reroute / Stop)
───────────────────────────────────────────────────────────────
Execution Protocol
Step 1: SCAN SYSTEM
Find all available skills:
# Check standard locations
ls -la ./.codex/skills/ 2>/dev/null || echo "No project skills"
ls -la ./skills/ 2>/dev/null || echo "No legacy project skills"
ls -la ./.claude/skills/ 2>/dev/null || echo "No legacy project-specific skills"
ls -la ~/.claude/skills/ 2>/dev/null || echo "No personal skills"
Step 2: ANALYZE QUERY
Understand the task:
- What is the real goal?
- What type of task is this?
- What skills are needed?
- What is the logical order?
Step 3: BUILD PLAN
Create execution plan:
- Select skills from scanned list
- Maximum 8 steps (prefer fewer)
- Each step = 1 skill (max 2 if tightly coupled)
- Order by dependencies
Step 4: SHOW PLAN
Present the plan to user:
- All steps with purposes
- Total count
- Ask for confirmation
Step 5: EXECUTE STEPS
For each step:
- Show step number and skill
- Execute the skill
- Show results
- Present execution status
- Ask to continue
Step 6: COMPLETE OR REBUILD
When all steps done:
- Show completion summary
- Or if user requests rebuild: create new plan
User Commands
| Command | Action |
|---|---|
| Yes / Continue / Next | Proceed to next step |
| Pause | Stop here, save progress |
| Reroute | Rebuild plan from current position |
| Skip | Skip current step, go to next |
| Back | Return to previous step |
| Stop | End execution, show summary |
Available Skills for Routing
The chain can include any of these skills:
| Skill | Use When |
|---|---|
goal-clarity-150 | Need to clarify requirements |
research-150 | Need internal investigation |
research-deep-150 | Need internal + external research |
impact-map-150 | Need to map dependencies |
deep-think-150 | Need quality reasoning |
proof-grade-150 | Need to verify critical claims |
action-plan-150 | Need to create detailed plan |
gated-exec-150 | Need controlled execution |
max-quality-150 | Need high-quality output |
refactor-150 | Need code restructuring |
coverage-70-tests | Need test coverage |
integrity-check-150 | Need final quality check |
tidy-up-150 | Need quick cleanup |
task-track-150 | Need status management |
lessons-learn | Need to capture learnings |
74-mid-session-save-150 | Need a mid-session checkpoint |
ask-ai-150 | Need external AI consultation |
Common Route Templates
🔧 Code Change Route
1. goal-clarity-150 → Clarify what to change
2. research-150 → Understand existing code
3. impact-map-150 → Map what's affected
4. action-plan-150 → Create implementation plan
5. gated-exec-150 → Implement with gates
6. coverage-70-tests → Add/verify tests
7. integrity-check-150 → Final quality check
🐛 Bug Fix Route
1. goal-clarity-150 → Clarify symptoms
2. research-150 → Investigate code/logs
3. deep-think-150 → Identify root cause
4. action-plan-150 → Plan the fix
5. gated-exec-150 → Apply fix
6. lessons-learn → Capture learnings
📝 Document Creation Route
1. goal-clarity-150 → Clarify document purpose
2. research-150 → Gather information
3. action-plan-150 → Create outline
4. max-quality-150 → Write document
5. integrity-check-150 → Review quality
🔍 Research Route
1. goal-clarity-150 → Clarify what to find
2. research-deep-150 → Internal + external research
3. proof-grade-150 → Verify findings
4. deep-think-150 → Synthesize conclusions
Output Format
End of Each Message (MANDATORY)
Every message during execution MUST end with this summary block:
───────────────────────────────────────────────────────────────
📊 EXECUTION STATUS
✅ Completed: [list of completed steps]
⚡ Current: Step N — [skill-name] — [status]
⏳ Remaining: [list of remaining steps]
📍 Progress: [progress bar] N/X steps (XX%)
🎯 Next: [next step description]
Continue? (Yes / Pause / Reroute / Stop)
───────────────────────────────────────────────────────────────
Plan Presentation Format
📋 **PLAN CREATED**
**Query:** [original user query]
**Goal:** [identified goal]
**Plan:**
| Step | Skill | Purpose |
|------|-------|---------|
| 1 | [skill] | [purpose] |
| 2 | [skill] | [purpose] |
| ... | ... | ... |
**Total:** N steps
Start execution? (Yes / Modify / Cancel)
Step Completion Format
✅ **Step N Complete**
**Skill:** [skill-name]
**Result:** [brief summary of what was accomplished]
**Artifacts:** [any outputs produced]
───────────────────────────────────────────────────────────────
📊 EXECUTION STATUS
[... standard status block ...]
───────────────────────────────────────────────────────────────
Execution Complete Format
🏁 **EXECUTION COMPLETE**
**Query:** [original query]
**Goal:** [goal] — ✅ ACHIEVED
**Summary:**
| Step | Skill | Status | Notes |
|------|-------|--------|-------|
| 1 | [skill] | ✅ | [note] |
| 2 | [skill] | ✅ | [note] |
| ... | ... | ... | ... |
**Total steps:** N
**Artifacts produced:** [list]
**What's next?** New task / Save session / End
Operational Rules
- ALWAYS SHOW STATUS: Every message ends with execution status block
- ONE STEP AT A TIME: Execute only the current step, then wait
- CONFIRMATION REQUIRED: Always ask before proceeding to next step
- REBUILD ON REQUEST: Create new plan when user asks
- CLEAR PROGRESS: Step counter visible at all times (Step N of X)
- SIMPLE LANGUAGE: Use clear, direct language
Failure Modes & Recovery
| Failure | Detection | Recovery |
|---|---|---|
| Wrong plan | User says "this isn't right" | Rebuild: create new plan |
| Step failed | Skill didn't produce expected output | Option to retry or skip |
| Lost context | User confusion | Show full status + plan |
| Too many steps | >8 steps | Simplify plan |
Examples
❌ Without Chain-of-Skills
User: "Add dark mode to the app"
AI: [Writes code immediately without planning]
Result: Incomplete, breaks other features, no tests
✅ With Chain-of-Skills
User: "Add dark mode to the app"
📋 PLAN CREATED
Query: Add dark mode to the app
Goal: Working dark mode toggle with proper styling
Plan:
| Step | Skill | Purpose |
|------|-------|---------|
| 1 | goal-clarity-150 | Clarify dark mode requirements |
| 2 | research-150 | Understand current styling system |
| 3 | impact-map-150 | Map all affected components |
| 4 | action-plan-150 | Create implementation plan |
| 5 | gated-exec-150 | Implement dark mode |
| 6 | coverage-70-tests | Add tests |
| 7 | integrity-check-150 | Final quality check |
Total: 7 steps
Start execution? (Yes / Modify / Cancel)
User: "Yes"
⚡ EXECUTING — Step 1 of 7
Current skill: goal-clarity-150
Purpose: Clarify dark mode requirements
[... skill execution ...]
───────────────────────────────────────────────────────────────
📊 EXECUTION STATUS
✅ Completed: (none yet)
⚡ Current: Step 1 — goal-clarity-150 — ✅ Complete
⏳ Remaining: Step 2-7
📍 Progress: ██░░░░░░░░░░░░░░ 1/7 steps (14%)
🎯 Next: Step 2 — research-150
Purpose: Understand current styling system
Continue? (Yes / Pause / Reroute / Stop)
───────────────────────────────────────────────────────────────
Relationship to Other Skills
- chain-flow-150 → More complex, full orchestration logic
- chain-of-skills → Simpler, user-friendly, step-by-step experience
- action-plan-150 → Creates plans within steps
- gated-exec-150 → Executes code within steps
Key difference: chain-of-skills is the user-facing execution experience, while other skills are the individual actions in each step.
Session Log Entry (MANDATORY)
After completing the chain, write to .sessions/SESSION_[date]-[name].md:
### [HH:MM] Chain-of-Skills 150 Complete
**Goal:** <original goal>
**Result:** <Success/Partial>
**Steps Executed:** <N>
**Artifacts:** <key outputs>
Remember: Scan skills → build plan → execute step by step → confirm each step. Simple, controlled, user in charge.
More Other skills
brand-guidelines
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
theme-factory
Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.
internal-comms
A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).
Fundamental design principles for writing maintainable, testable Python code.
responsive-design
Modern responsive layouts using container queries, fluid typography, CSS Grid, and mobile-first strategies.
Generate Playwright tests from scenarios using interactive browser exploration and validation.
Automated website exploration and test case generation using Playwright.
Modern React state management with Redux Toolkit, Zustand, Jotai, and React Query for every state category.
gdpr-data-handling
GDPR-compliant data handling with consent management, data subject rights, and privacy controls.
Python error handling patterns including input validation, exception hierarchies, and partial failure handling. Use when implementing validation logic, designing exception strategies, handling batch processing failures, or building robust APIs.
refactor-plan
Detailed multi-file refactor planning with safety checks and rollback guidance.
