meow:lazy-agent-loader
Load agent definitions on-demand to reduce context usage — only loads full agent when scoring indicates it's needed.
What This Skill Does
Instead of loading all 13 agent definitions at startup (~26,000 tokens), this skill loads only the agent index (~1,000 tokens) and fetches full definitions when a specific agent scores high enough to be activated. Saves 90%+ of context for single-agent tasks.
Core Capabilities
- Agent index — Compact keyword-based index for all 13 agents (~1,000 tokens)
- Scoring-based loading — Score ≥80: load full definition. 50-79: summary only. <50: skip.
- Session cache — Once loaded, agent stays cached for the session
Usage
Automatic — used by meow:agent-detector for optimized loading.
Skill Details
Phase: 0
Used by: orchestrator agent
Gotchas
- Agent loaded too late missing context: Lazy loading skips context that was available at session start → Pre-load agents that need session-start context (orchestrator, analyst)
- Cache serving stale agent definition: Agent file updated but cached version used → Invalidate cache on file mtime change
Related
meow:agent-detector— Triggers lazy loading based on scoring