01 / how it works

Static analysis for the file you already wrote.

ccmd does not run your agent. It reads your config file, scores it against a transparent rubric, and tells you which lines are silently bloating your context. The free tier never sends the file anywhere; every check runs in your browser tab.

  1. 01

    Detect file type

    We sniff for CLAUDE.md, AGENTS.md, .cursorrules, or .grokrules from the header tokens, language markers, and structural shape. Same rubric runs against all four. You do not pick a mode.

  2. 02

    Token math

    Total tokens, lines that fire on every turn, percent of context budget at standard Anthropic / OpenAI / xAI rates, projected dollar cost across a 30-turn long-running session at Opus 4.7 rates. The free tier uses char/4 estimation; the paid tier uses the real Anthropic tokenizer.

  3. 03

    Karpathy-12 rubric

    Twelve rules, each a regex-and-heuristic check: Think-Before-Coding, Simplicity-First, Surgical-Changes, Goal-Driven-Execution, Avoid-Silent-Assumptions, No-Orthogonal-Damage, Tests-as-Truth, Concise-Output, Stack-Awareness, Tool-Preference, Failure-Mode-Coverage, Self-Improvement-Loop. Community measurement: 41% → 11% mistake-rate drop when all twelve are present.

  4. 04

    Structural lints

    Per-line scan for token bloat (lines over 28 words), vague terms (appropriate, properly, carefully), aspirational rules (always / never with no escape clause), missing-why prohibitions, exact duplicates, contradictions, and cache-busting timestamps near the top of the file.

  5. 05

    Optimized copy

    The 'Copy optimized version' button strips the safest cuts: exact duplicates and cache-bust lines. Paid tier offers full structural rewrite with a side-by-side diff and PR description.

  6. 06

    Paid tier additions

    Session-trace analysis (which rules actually fired in your last 100 Claude Code sessions), continuous drift monitoring via GitHub webhook, weekly drift email, PR diff comments on CLAUDE.md changes, per-engineer cost attribution, team-shared dashboard.