measured-coding-context-audit

Estimates where a coding agent's context window is being spent and audits the usual waste: duplicate skills, rules repeated between global and project files, oversized instruction files, redundant MCP servers, dirty git status, and very long sessions — then reports savings ranked by impact. Use when the user says "context checkup", "how big is my context", "reduce context usage", "my agent feels sluggish", or "audit my Claude config". Not for measuring API billing or costs, tuning model parameters, or compressing LLM model weights.

Version

v1.0.1

Size

10 KB

Updated

2026-10-06

Your client verifies the signature before reading. This explicit manual read is available on every plan. Free automatically loads five chosen skills; paid plans automatically load relevant skills across the library.