• GitHub pattern → EXECUTE Repository flow
• YouTube pattern → EXECUTE Video transcript flow
• skills.sh pattern → EXECUTE Skill catalog flow
• skill:// pattern → EXECUTE Skill resolution flow
• Article pattern → EXECUTE Web scraping flow </decision_tree>

1. EXECUTE: git clone <url> ~/lev/workshop/intake/<repo_name>
2. VERIFY: Repository cloned successfully
3. SAVE STATUS: "Repository ready for analysis"

1. Route ALL video/media URLs through ~/digital/homie/yt/
2. EXECUTE PRIMARY: cd ~/digital/homie && python yt/cli.py -t "<url>" --wait
3. IF PRIMARY FAILS: EXECUTE FALLBACK: python yt/yt.py -t "<url>" --wait
4. IF BOTH FAIL: mcp__fetch-mcp__fetch_youtube_transcript (YouTube only)
5. SAVE TRANSCRIPT: Create ~/lev/workshop/intake/transcript-{video_id}.txt with content
6. VERIFY: Transcript file exists and contains content NOTE: Supports 100+ platforms via yt-dlp. After transcription: "Where should this content go?"

1. EXECUTE PRIMARY: cd ~/cb && python scraping_orchestrator.py <url>
2. IF PRIMARY FAILS: EXECUTE FALLBACK: mcp__firecrawl__firecrawl_scrape
3. SAVE CONTENT: Create ~/lev/workshop/intake/content-{domain}.txt with scraped content
4. VERIFY: Content file exists and contains scraped data

1. Detect: Is this a skill/skills.sh URL?
2. ROUTE: Hand off to skill-builder for full lifecycle
   • skill-builder will: validate → prior art check → install → score → lifecycle
3. RETURN: skill-builder reports back with install status NOTE: lev-intake does NOT install skills directly. skill-builder owns the full lifecycle.

1. RESOLVE: Search ~/.agents/skills/ and ~/.agents/skills-db/ for matching skill
2. IF NOT FOUND: Search skills.sh marketplace
3. INSTALL: Copy/clone skill to ~/.agents/skills-db/_workshop/{name}/
4. VERIFY: SKILL.md exists at destination
5. SAVE STATUS: "Skill acquired, ready for analysis" </content_routing>

1. EXECUTE: cat ~/lev/workshop/intake.md
2. FOLLOW: The complete analysis framework loaded from that file
3. COMPLETE: All Phase 2 and Phase 3 steps as defined in workshop/intake.md

• Cache scanning for existing capabilities
• Lev system overlap detection
• LLM-first evaluation criteria
• Strategic tier classification
• Interactive ADR creation process
• Post-processing decisions

<progress_template> INTAKE PROGRESS:

• Show user what's in ~/lev/workshop/intake/
• Show pending analysis items
• Ask: "This is [content type]. Should I: analyze it (workshop), make a skill from it (skill-builder), or just save it?"

• Skills → skill-builder (full lifecycle)
• Repos/articles → workshop/intake.md (Phases 2-3 analysis)
• Video/media → ~/digital/homie/yt/ pipeline → then route the output
• Unknown → show dashboard and ask user </final_reminder>

• Lev Master Router (

  lev/SKILL.md

  ) - Routes all lev-* skills Parent skill that dispatches to this skill based on keywords/context

• Role definition - Clarifies operating scope and prevents ambiguous execution.
• Context enrichment - Captures required inputs before actions.
• Output structuring - Standardizes deliverables for consistent reuse.
• Step-by-step workflow - Reduces errors by making execution order explicit.
• Edge-case handling - Documents safe fallbacks when assumptions fail.

• Required: clear user intent and relevant context for this skill.
• Preferred: repository/project constraints, existing artifacts, and success criteria.
• If context is missing, ask focused questions before proceeding.

• Provide structured, actionable outputs aligned to this skill's existing format.
• Include assumptions and next steps when appropriate.
• Preserve compatibility with existing sections and related skills.

• If prerequisites are missing, provide a minimal safe path and request missing inputs.
• If scope is ambiguous, narrow to the highest-confidence sub-task.
• If a requested action conflicts with existing constraints, explain and offer compliant alternatives.