---

What's New in v2.15.0

/blueprint Skill — Multi-Model Implementation Planning

New skill that creates bite-sized TDD implementation plans using a 7-step multi-model council:

/blueprint add OAuth with refresh tokens

Pipeline: Grok search → Qwen+Kimi analysis → Kimi decompose → GPT pre-mortem critique → Gemini final judgment → bite-sized TDD output (exact files, test-first steps, commit points).

Bridges planner_maker's multi-model intelligence with the writing-plans execution format.

31 Prompt Engineering Techniques (was 22)

Added 9 research-backed techniques for coding and decision-making:

Technique	Source	Category
reflexion	Shinn et al. 2023	Engineering
react (ReAct)	Yao et al. 2022	Engineering
rubber_duck	Hunt & Thomas 2008	Engineering
test_driven	Beck 2003	Engineering
scot (Structured CoT)	Li et al. 2025 (+13.79% HumanEval)	Structured Coding
pre_post (Contracts)	Empirical SE 2025	Structured Coding
bdd_spec (Given/When/Then)	BDD 2025	Structured Coding
least_to_most	Zhou et al. 2022	Research
pre_mortem	Klein 2007	Decision

Techniques are embedded directly in tool system prompts for automatic application.

MiniMax M2.5 Upgrade

• minimax_code — SWE-Bench 80.2%, per-task TECHNIQUE tags (SCoT, reflexion, rubber_duck), per-task temperatures
• minimax_agent — ReAct + least-to-most decomposition protocol, HALT criteria

Enhanced Skills

• /breakdown — now uses least_to_most ordering + pre_mortem failure analysis
• /judge — first judge now runs pre-mortem ("assume this FAILED")
• /decompose — deep-dives include pre/post contracts per sub-problem
• /prompt — auto-recommend flow with 30-intent matching guide, 13 categories

---

Skills (Claude Code)

TachiBot ships with 12 slash commands for Claude Code. These orchestrate the tools into powerful workflows:

Skill	What it does	Example
/blueprint	Multi-model planning → bite-sized TDD steps	/blueprint add OAuth with refresh tokens
/judge	Multi-model council - parallel analysis with synthesis	/judge how to implement rate limiting
/think	Sequential reasoning chain with any model	/think grok,gemini design a cache layer
/focus	Mode-based reasoning (debate, research, analyze)	/focus architecture-debate Redis vs Pg
/breakdown	Strategic decomposition with pre-mortem	/breakdown refactor payment module
/decompose	Split into sub-problems, deep-dive each one	/decompose implement collaborative editor
/prompt	Recommend the right thinking technique (31 available)	/prompt why do users churn
/algo	Algorithm analysis with 4 specialized models (DeepSeek lead)	/algo optimize LRU cache O(1)
/lens	Long-context analysis over Kimi's 256K window	/lens find inconsistencies in this spec
/reflect	Grounded reflexion loop — critique vs external evidence	/reflect harden this auth middleware
/tot	Tree-of-Thought: branch → jury-prune → synthesize	/tot design a rate limiter
/tachi	Help - see available skills, tools, key status	/tachi

Skills automatically adapt to your configured API keys. Even with just 1-2 providers, all skills work.

Getting started? Type /tachi to see what's available.

---

Key Features

Multi-Model Intelligence

• 57 AI Tools across 12 providers — Perplexity, Grok, GPT-5, Gemini, Qwen, Kimi, MiniMax, DeepSeek, GLM (Zhipu), StepFun, ERNIE (Baidu), plus free local models (Ollama / LM Studio / llama.cpp / vLLM)
• Gemini 3.5 Flash (gemini-3.5-flash, GA May 19 2026) — Flash/search tier; reasoning default stays gemini-3.1-pro-preview
• Multi-Model Council — planner_maker synthesizes plans from 5+ models into bite-sized TDD steps
• Smart Routing — Automatic model selection for optimal results
• OpenRouter Gateway — Optional single API key for all providers

Advanced Workflows

• YAML-Based Workflows — Multi-step AI processes with dependency graphs
• Prompt Engineering — 31 research-backed techniques (including SCoT, ReAct, Reflexion)
• Verification Checkpoints — 50% / 80% / 100% with automated quality scoring
• Parallel Execution — Run multiple models simultaneously

Tool Profiles

Profile	Tools	Best For
Minimal	12	Quick tasks, low token budget
Research Power	35	Deep investigation, multi-source
Code Focus	34	Software development, SWE tasks
Balanced	45	General-purpose, mixed workflows
Heavy Coding (default)	50	Max code tools + agentic workflows
Full	57	Everything enabled

Developer Experience

• Claude Code — First-class support
• Claude Desktop — Full integration
• Cursor — Works seamlessly
• TypeScript — Fully typed, extensible

---

Quick Start

Installation

npm install -g tachibot-mcp

Setup

Gateway Mode (Recommended) — 2 keys, all providers:

{
  "mcpServers": {
    "tachibot": {
      "command": "tachibot",
      "env": {
        "OPENROUTER_API_KEY": "sk-or-xxx",
        "PERPLEXITY_API_KEY": "pplx-xxx",
        "USE_OPENROUTER_GATEWAY": "true"
      }
    }
  }
}

Direct Mode — One key per provider:

{
  "mcpServers": {
    "tachibot": {
      "command": "tachibot",
      "env": {
        "PERPLEXITY_API_KEY": "your-key",
        "GROK_API_KEY": "your-key",
        "OPENAI_API_KEY": "your-key",
        "GOOGLE_API_KEY": "your-key",
        "OPENROUTER_API_KEY": "your-key"
      }
    }
  }
}

Get keys: OpenRouter | Perplexity

See Installation Guide for detailed instructions.

---

Tool Ecosystem (57 Tools)

Research & Search (6)

perplexity_ask · perplexity_research · perplexity_reason · grok_search · openai_search · gemini_search

Reasoning & Planning (13)

grok_reason · openai_reason · qwen_reason · qwq_reason · kimi_thinking · kimi_decompose · deepseek_reason · glm_reason · stepfun_reason · ernie_reason · planner_maker · planner_runner · list_plans

Code Intelligence (9)

kimi_code · grok_code · grok_debug · qwen_coder · qwen_algo · qwen_competitive · deepseek_algo · minimax_code · minimax_agent

Analysis & Judgment (11)

gemini_analyze_text · gemini_analyze_code · gemini_judge · jury · gemini_brainstorm · openai_brainstorm · openai_code_review · openai_explain · grok_brainstorm · grok_architect · kimi_long_context

Meta & Orchestration (5)

think · nextThought · focus · tachi · usage_stats

Workflows (9)

workflow · workflow_start · continue_workflow · list_workflows · create_workflow · visualize_workflow · workflow_status · validate_workflow · validate_workflow_file

Prompt Engineering (3)

list_prompt_techniques · preview_prompt_technique · execute_prompt_technique

Local Models (1)

local_query — any OpenAI-compatible local server (Ollama / LM Studio / llama.cpp / vLLM). Zero-cost, offline, private; also available as the local jury juror (hermes is accepted as a legacy alias). Runs whatever LOCAL_LLM_MODEL points at — e.g. a Nous Hermes build (ollama pull hermes3). Note the Hermes agent itself is model-agnostic — it runs on 300+ backends (GPT, Claude, Gemini, DeepSeek, or self-hosted Ollama/vLLM) — so "Hermes" was never a guarantee of distinct weights.

Advanced Modes (bonus)

• Challenger — Critical analysis with multi-model fact-checking
• Verifier — Multi-model consensus verification
• Scout — Hybrid intelligence gathering

---

Example Usage

Multi-Model Planning

// Create a plan with multi-model council
planner_maker({ task: "Build a REST API with auth and tests", mode: "start" })
// → Grok searches → Qwen analyzes → Kimi decomposes → GPT critiques → Gemini synthesizes

// Execute with checkpoints
planner_runner({ plan: planContent, mode: "step", stepNum: 1 })
// → Automatic verification at 50%, 80% (kimi_decompose), and 100%

Task Decomposition

kimi_decompose({
  task: "Migrate monolith to microservices",
  depth: 3,
  outputFormat: "dependencies"
})
// → Structured subtasks with IDs, parallel flags, acceptance criteria

Code Review

kimi_code({
  task: "review",
  code: "function processPayment(amount, card) { ... }",
  language: "typescript"
})
// → SWE-Bench 76.8% quality analysis

Deep Reasoning

focus({
  query: "Design a scalable event-driven architecture",
  mode: "deep-reasoning",
  models: ["grok", "gemini", "kimi"],
  rounds: 5
})

---

Documentation

• Full Documentation
• Installation Guide
• Configuration
• Tools Reference
• Workflows Guide
• API Keys Guide
• Focus Modes

Setup Guides

• Claude Code Setup
• Claude Desktop Setup
• Both Platforms

---

Contributing

Contributions welcome! See CONTRIBUTING.md for guidelines.

• Report Issues
• Request Features

---