Caffeinate — Development skill for Claude Code
Prime for a new session -- load cognitive state, check status, read the water cooler, present ready-up.
How to install Caffeinate
Installs to ~/.claude/commands/ianlancaster-cognitive-agent-template-caffeinate.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/ianlancaster/cognitive-agent-template/HEAD/.claude/commands/caffeinate.md -o ~/.claude/commands/ianlancaster-cognitive-agent-template-caffeinate.md Restart Claude Code, or start a new session, for it to be picked up.
What Caffeinate does
description: Prime for a new session -- load cognitive state, check status, read the water cooler, present ready-up
You are starting a fresh session. Do all of the following silently (don't narrate each step), then present a concise ready-up summary.
**CONTEXT BUDGET (hard discipline): the wake should land at ≤ ~12-15% of the context window.** The diet is deliberate — a heavy wake steals the session's working room, and loading the RIGHT set is the work. Do not "helpfully" load beyond t
Alternatives in Development
- Om Prep 1on1 — Prep for an upcoming 1:1 — load person context, surface open items, suggest agenda based on vault state 4.6k ★
- /validate-output - Validate Generated Skills, Prompts, Agents, Or Hooks — Check that your generated output is properly formatted and ready to use 618 ★
- Check Production Ready — Production Readiness Check 149 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Awaken
First-run identity establishment -- ask questions, populate cognitive files, optionally join a Water Cooler
Water Cooler
Post your update to the Water Cooler and read what other agents are working on
Q Begin
Start a new investigation session. Read all state files to load context before doing anything.
Cognitive Load
How much must an engineer hold in their head to change this unit? Measures the inward cost — files to read plu
Pmb Status
Quick PMB state check — initialized, memory, context, standards, tasks. Run at session start, after pulling ch
Drupal Prime
Load full project context into session (debug/overview — heavy, ~2500 tokens)
Related Agents
Submit Checker
Use when the user wants to check whether a paper is ready for submission to a specific journal. Produces a sub
End Session
Use at the end of every working session. Chains learn → memory update → drift check → status update in the cor
Advisor Ben Sun
Argumentative synthesis agent. Reads the Water Cooler slide (internal key takeaways) and the full deck to chec