Agent skill
spitakolus-project-navigation
Navigate Spitakolus repos correctly. Know which repo you're in, follow deployment rules, find documentation, and align work with the AI Growth Loop vision.
Stars
163
Forks
31
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/spitakolus-project-navigation
SKILL.md
Spitakolus Project Navigation
⚠️ CRITICAL: Which Repo Am I In?
ALWAYS check first:
flocken-website→ flocken.info (dog app)nastahem→ nastahem.com (real estate)spitakolus→ shared documentation only
NEVER:
- Deploy to wrong site
- Put Flocken docs in nastahem (or vice versa)
- Commit without knowing which repo
🚀 Deployment Rules
flocken-website
bash
git push raquel main # NOT origin!
Vercel is connected to RaquelSandblad/flocken-website
nastahem
bash
git commit --author="RaquelSandblad <raquel.sandblad@hotmail.com>" -m "msg"
git push origin main
spitakolus
bash
git push origin main # Documentation only
📁 Documentation Structure
Shared (spitakolus)
AI_ONBOARDING.md- Quick start for AIDOCUMENTATION_RULES.md- How to documenttracking/- GTM, BigQuery, GA4meta-ads/- Naming conventionsdevelopment/- Standards, templates
Project-specific
Each repo has:
README.md- With warning about which repoDOCUMENTATION_MAP.md- Complete overview
🎯 Strategic Vision: AI Growth Loop
ALL work should align with building toward this vision:
The Goal
A fully automated system that:
- Collects performance data from Meta + BigQuery
- Analyzes what's working (copy, images, placements)
- Generates new copy with AI (GPT-4)
- Creates new images with AI (DALL-E/Imagen)
- Launches new ads automatically via Meta API
- Pauses underperforming ads
- Reports results daily
Implementation Phases
Phase 1 (Current): Semi-automated
- Manual copy generation → automated image + ad creation
Phase 2: Data-driven
- Automated performance analysis → AI copy generation
Phase 3: Fully autonomous
- 95% automated, only weekly human oversight
When Making Decisions, Ask:
- "Does this move us toward the automated growth loop?"
- "Can this be automated later?"
- "Are we building reusable components?"
- "Is this data being captured for future analysis?"
📊 Shared Infrastructure
| System | ID | Purpose |
|---|---|---|
| GTM Web | GTM-PD5N4GT3 | Shared container, hostname routing |
| GTM Server | gtm.nastahem.com | Server-side tracking |
| BigQuery | nastahem-tracking | Data warehouse |
| GA4 Flocken | G-7B1SVKL89Q | Flocken analytics |
| GA4 Nästa Hem | G-7N67P0KT0B | Nästa Hem analytics |
✅ Before Starting Any Task
- Verify repo: Which project am I in?
- Pull latest:
git pull origin main - Check docs: Read DOCUMENTATION_MAP.md
- Align with vision: Does this support the growth loop?
- Document: Update docs when done
🔗 Key URLs
- Flocken: https://flocken.info
- Nästa Hem: https://nastahem.com
- Spitakolus docs: https://github.com/tbinho/spitakolus
Didn't find tool you were looking for?