README Compression section.

proactively before requests hit upstream providers. This means your token savings happen transparently — no changes needed to your workflow.

---

Best for: Always-on usage, safety-critical workflows.

Caveman — removes filler words and verbose phrasing while preserving meaning:

Best for: Daily coding workflows, cost-conscious teams.

Message Aging — older messages get progressively compressed
• Tool Result Summarization — long tool outputs replaced with summaries
• Structural Integrity Guards — ensures + pairs stay consistent
• Context Window Awareness — respects per-model token limits

Best for: Extended debugging sessions, large codebases.

Heuristic Pruning — removes messages below relevance threshold
• Code Block Thinning — compresses repetitive code examples
• Binary Search Truncation — finds optimal cut point for context window

Best for: When you're hitting context limits repeatedly.

, , , test runners, TypeScript/Vite/Webpack builds, ESLint/Biome/Prettier, npm audit/installs, Docker logs, infra output, and generic shell output

Best for: Agent sessions with shell, build, test, git, grep, and file-output transcripts.

RTK -> Caveman

Best for: Mixed context with large tool logs plus human instructions or assistant summaries.

---

	fewer output tokens, benchmark average output savings, range, and input compression tool
	command-output savings; sample session tokens, or saved ()

RTK -> Caveman

combined = 1 - (1 - RTK savings) * (1 - Caveman input savings)
average  = 1 - (1 - 0.80) * (1 - 0.46) = 89.2%
range    = 1 - (1 - 0.60..0.90) * (1 - 0.46) = 78.4-94.6%

number applies when both RTK and Caveman can reduce the same input/context payload. Caveman response output mode is separate: when enabled, use Caveman's own output savings ( average, headline, range). Total billing savings depend on your prompt/output mix.

---

---

:

Caveman — mode selection, language packs, preview, and global defaults
• RTK — command-filter preview, RTK safety settings, and filter catalog
• Compression Combos — named engine pipelines assigned to routing combos
• Auto-Trigger Threshold — automatically engage compression when token count exceeds threshold

, assign a compression combo to a routing combo:

Combo: "free-forever"
  Compression Combo: "coding-agent-stack"
  Pipeline: RTK -> Caveman
  Targets:
    1. gc/gemini-3-flash
    2. if/kimi-k2-thinking

# Get compression settings
curl http://localhost:20128/api/settings/compression

# Update compression settings
curl -X PUT http://localhost:20128/api/settings/compression \
  -H "Content-Type: application/json" \
  -d '{"defaultMode":"stacked","autoTriggerMode":"stacked","autoTriggerTokens":32000}'

# Preview a specific RTK/stacked payload
curl -X POST http://localhost:20128/api/compression/preview \
  -H "Content-Type: application/json" \
  -d '{"mode":"rtk","messages":[{"role":"tool","content":"npm test output here"}]}'

# List RTK filter packs
curl http://localhost:20128/api/context/rtk/filters

# Test RTK directly with optional command metadata
curl -X POST http://localhost:20128/api/context/rtk/test \
  -H "Content-Type: application/json" \
  -d '{"command":"npm test","text":"FAIL tests/example.test.ts\nError: boom"}'

---

always preserves:

---

{
  "originalTokens": 47200,
  "compressedTokens": 40120,
  "savingsPercent": 15.0,
  "techniquesUsed": ["collapseWhitespace", "dedupSystemPrompt"],
  "mode": "lite",
  "engine": "caveman",
  "compressionComboId": "coding-agent-stack",
  "durationMs": 0.8,
  "rtkRawOutputPointers": []
}

---

---

Caveman by JuliusBrussee (⭐ 51K+) — the viral "why use many token when few token do trick" project. Caveman reports fewer output tokens, benchmark average output savings, a output range, and a input-compression tool.

RTK - Rust Token Killer by RTK AI — the high-performance command-output compression project for terminal, build, test, git, and tool-output filtering. RTK reports savings, with its README sample session showing saved.

---

Environment Config — Compression environment variables
• Architecture Guide — Compression pipeline internals
• User Guide — Getting started with compression
• RTK Compression — RTK filters, trust model, verify gate, raw-output recovery
• Compression Engines — Caveman, RTK, stacked, APIs, MCP, dashboard
• Compression Rules Format — JSON rule-pack format
• Compression Language Packs — Language-specific Caveman rules