Agent Token Optimizer Report
sample-product-repo ยท Example data for GitHub Pages preview
This is a safe sample report. It uses rounded fixture values and generic file paths, not a real private repository scan.
Total tokens1,248,000
Recommended context92,000
Reducible tokens1,156,000
Reduction rate93%
AI Readiness
82
Token Efficiency
71
Context Hygiene
76
Model Routing
68
Security Risk
91
Files scanned428
Excluded candidates63
Generated/lock/binary files31
Secret-risk files0
What To Fix Next
- high Exclude generated, binary, and lock files from default AI context. -760,000 tokens
- medium Split large Markdown documentation into task-specific `docs/ai/` files.
- medium Add context presets for focused, feature, and architecture tasks.
Top Token Files
| File | Tokens | Kind | Action |
|---|---|---|---|
| docs/product/long-spec.md | 186,000 | docs | split |
| package-lock.json | 154,000 | lockfile | exclude |
| dist/app.bundle.js | 142,000 | generated | exclude |
| AGENTS.md | 18,200 | agent-instructions | route |
Top Token Directories
| Directory | Tokens | Files |
|---|---|---|
| docs | 420,000 | 88 |
| dist | 260,000 | 24 |
| src | 198,000 | 126 |
| tests | 74,000 | 52 |
Model Routing
| Task | Tier | Context | Max tokens |
|---|---|---|---|
| Small docs edit | low | focused | 32,000 |
| Feature implementation | medium | feature | 96,000 |
| Architecture review | high | architecture | 160,000 |