Tok Opt — Development skill for Claude Code
Activate the unified Token Optimizer suite to slash token consumption by 50-70% across high-end models (caveman + investigate-first + surgical-patch + session-compaction).
How to install Tok Opt
Installs to ~/.claude/skills/nikhils1229-token-optimizer-skills-tok-opt/SKILL.md
mkdir -p ~/.claude/skills/nikhils1229-token-optimizer-skills-tok-opt && curl -fsSL https://raw.githubusercontent.com/nikhils1229/token-optimizer-skills/HEAD/commands/tok-opt.md -o ~/.claude/skills/nikhils1229-token-optimizer-skills-tok-opt/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Tok Opt does
description: Activate the unified Token Optimizer suite to slash token consumption by 50-70% across high-end models (caveman + investigate-first + surgical-patch + session-compaction)
Token Optimizer Master Command
When the user runs `/token-optimizer` or `/tok-opt`, or asks to "optimize tokens", "minimize tokens", or "use caveman mode":
- Activate the Token Optimizer Protocol (
~/.gemini/config/skills/token-optimizer/SKILL.md):- Output Density: Switch to Caveman compr
Alternatives in Development
- GenericAgent — Self-evolving agent: grows skill tree from 3.3K-line seed, achieving full system control with 6x less token co 6.1k ★
- Patch 2.3k ★
- Investigate — Investigation Skill 1.9k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Token Optimizer
Activate the unified Token Optimizer suite to slash token consumption by 50-70% across high-end models (cavema
Output Compress
Tiered, opt-in output compression skill with a deterministic fidelity gate (Claude Code Agent Skill + Codex AG
Jmunch Console
Local, opt-in control-plane GUI for the jMunch suite (config, index/watcher health, savings/ROI, sessions, age
Ensoul
Activate Claudicle soul identity in this session. Creates a per-session marker so soul.md, soul state, and ses
Warden Select
Measure pending token-warden candidate rules for an agent on the golden suite, evict or activate them, and rec
Kaggle Tpu Lab
Qwen3.8-27B (bf16) on a free Kaggle TPU: OpenAI-compatible endpoint, 262k context, ~130 tok/s, works with Clau
Related Agents
Deep Research Patcher
Applies critic findings to a nav-deep-research draft as surgical Edit hunks and records a patch log. Tool-lock
Suede Graph Flo XR Patch Author
Read-only Suede Graph Flo XR worker that inspects selected files and returns structured unified diffs without
Patch Engineer
Drafts a minimal idiomatic unified diff that closes exactly the cited bound, then verifies it by re-running th