Load and review Emergent Learning Framework context, institutional knowledge, golden rules, and recent session history. Runs the checkin workflow interactively with banner, context loading, and dashboard/multi-model prompts.
Interactive workflow to load the building context before starting work.
The /checkin command:
/checkin
The checkin command is simple - just type /checkin to load framework context and prepare your session.
This skill runs the new Python-based orchestrator:
python ~/.claude/emergent-learning/elf.py checkin
# OR directly:
python ~/.claude/emergent-learning/src/query/checkin.py
The orchestrator is a complete 8-step workflow:
Show ELF ASCII art immediately
Query the learning framework
Parse and format context for readability
Spawn async haiku agent to summarize recent work
Ask user if they want to start the dashboard
Interactive prompt to select your active AI model
ELF_MODEL environment variableCheck for pending CEO decisions in ceo-inbox/
Print completion message
✅ Banner First - Displayed before any prompts, not after
✅ One-Time Prompts - Dashboard and model selection appear only on first checkin
✅ State Tracking - Uses ~/.claude/.elf_checkin_state to track conversation state
✅ Model Persistence - Selection stored in ELF_MODEL environment variable
✅ Structured Workflow - All 8 steps executed in proper sequence
✅ Context Parsing - Query output properly formatted for display
Start ELF Dashboard?
The dashboard provides metrics, model routing, and system health.
Start Dashboard? [Y/n]:
Select Your Active Model
Available models:
(c)laude - Orchestrator, backend, architecture (active)
(g)emini - Frontend, React, large codebases (1M context)
(o)dex - Graphics, debugging, precision (128K context)
(s)kip - Use current model
Select [c/g/o/s]:
ELF_MODEL environment variableThe checkin workflow is your gateway to the building's knowledge:
Running checkin at the start of each session ensures you're working with current institutional knowledge.