tachibot-mcp 2.23.0 → 2.23.3

package/CHANGELOG.md

@@ -5,6 +5,24 @@ All notable changes to TachiBot MCP will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [2.23.3] - 2026-06-17
+
+### Changed
+- **GLM-5.1 → GLM-5.2** (`z-ai/glm-5.2`, released Jun 13 on OpenRouter, verified via grok_search + web): Zhipu's open-weights flagship for long-horizon tasks — usable 1M-token context (up from agentic-tier 5.1), two thinking-effort levels (High/Max). SWE-Bench Pro 62.1 (up from 58.4), Terminal-Bench 2.1 81.0 (up from 62.0); beats GPT-5.5 on several long-horizon coding benchmarks at ~1/6 the cost. List price $1.40/$4.40 per M (262K max output). Powers `glm_reason` and the `glm` juror. GLM-5.1 kept as quota fallback via `MODEL_FALLBACKS` (GLM-5.2 → GLM-5.1 → GLM-5); the `glm` substring already routes 5.2 to the 600s reasoning timeout.
+
+## [2.23.2] - 2026-06-15
+
+### Changed
+- **Kimi K2.6 → K2.7-Code** (`moonshotai/kimi-k2.7-code`, released Jun 12 on OpenRouter, verified via grok_search): coding-specialized variant built on K2.6 — +21.8% on Kimi Code Bench v2 with lower token use, 262K context, native multimodal, always-thinking mode. List price $0.75/$3.50 per M. Now powers all Kimi tools — `kimi_thinking`, `kimi_code`, `kimi_decompose`, `kimi_long_context` (`openrouter-tools.ts`) and the `kimi` juror (`jury-tool.ts`). K2.6 kept as quota fallback via `MODEL_FALLBACKS` (K2.7-Code → K2.6 → K2-Thinking); the `kimi` substring already routes K2.7-Code to the 600s reasoning timeout (regression test added).
+
+### Notes
+- No general-purpose Kimi K2.7 exists yet — K2.7-Code is the only post-K2.6 release. The coding-tuned model now also serves the reasoning/long-context tools and the kimi juror.
+
+## [2.23.1] - 2026-06-11
+
+### Changed
+- **MiniMax M2.7 → M3** (`minimax/minimax-m3`, released May 31 on OpenRouter, API-verified via grok_search): 1M-token context (up from 200K), MSA sparse attention (~1/20 compute at 1M ctx vs prior gen), native multimodal, tuned for long-horizon agentic work. Same promo pricing as M2.7 ($0.30/$1.20 per M). Powers `minimax_code`, `minimax_agent`, and the `minimax` juror. M2.7 kept as quota fallback via `MODEL_FALLBACKS`.
+
 ## [2.23.0] - 2026-06-11
 
 ### Changed

package/README.md

@@ -4,7 +4,7 @@
 
 ### Multi-Model AI Orchestration Platform
 
-[![Version](https://img.shields.io/badge/version-2.15.2-blue.svg)](https://www.npmjs.com/package/tachibot-mcp)
+[![Version](https://img.shields.io/badge/version-2.23.2-blue.svg)](https://www.npmjs.com/package/tachibot-mcp)
 [![Tools](https://img.shields.io/badge/tools-57_active-brightgreen.svg)](#-tool-ecosystem-57-tools)
 [![License](https://img.shields.io/badge/license-AGPL--3.0-green.svg)](LICENSE)
 [![Node](https://img.shields.io/badge/node-%3E%3D18.0.0-brightgreen.svg)](https://nodejs.org)
@@ -12,7 +12,7 @@
 
 **57 AI tools. 12 providers. One protocol.**
 
-Orchestrate Perplexity, Grok, GPT-5, Gemini, Qwen, Kimi K2.5, and MiniMax M2.5
+Orchestrate Perplexity, Grok, GPT-5.5, Gemini, Qwen, Kimi K2.7-Code, and MiniMax M3
 from Claude Code, Claude Desktop, Cursor, or any MCP client.
 
 [Get Started](#-quick-start) · [View Tools](#-tool-ecosystem-57-tools) · [Documentation](https://tachibot.com/docs)

package/dist/src/config/model-constants.js

@@ -87,17 +87,20 @@ export const GROK_MODELS = {
     _3: "grok-3", // Legacy with search: 256K→2M
 };
 // Kimi Models (Moonshot AI via OpenRouter)
-// K2.6 released Apr 20, 2026 - 1T MoE, leads SWE-bench Pro for long-horizon coding
-// K2.5 released Jan 27, 2026 - kept as fallback
+// K2.7-Code released Jun 12, 2026 - coding-specialized (built on K2.6), +21.8% on Kimi Code Bench v2, lower token use, 262K ctx, multimodal, always-thinking
+// K2.6 released Apr 20, 2026 - 1T MoE, leads SWE-bench Pro for long-horizon coding (fallback)
+// K2.5 released Jan 27, 2026 - kept as legacy
 export const KIMI_MODELS = {
     K2_THINKING: "moonshotai/kimi-k2-thinking", // 1T MoE, 32B active - agentic reasoning (256k context)
-    K2_6: "moonshotai/kimi-k2.6", // CURRENT: 1T MoE, SWE-Pro leader, ~$0.74/$4.65
-    K2_5: "moonshotai/kimi-k2.5", // Previous: multimodal + agent swarm (fallback)
+    K2_7_CODE: "moonshotai/kimi-k2.7-code", // CURRENT (Jun 12, 2026): coding-specialized, +21.8% Code Bench v2, 262K ctx, $0.75/$3.50
+    K2_6: "moonshotai/kimi-k2.6", // Previous (Apr 2026): 1T MoE, SWE-Pro leader (fallback)
+    K2_5: "moonshotai/kimi-k2.5", // Legacy: multimodal + agent swarm
 };
 // MiniMax Models (MiniMax via OpenRouter)
-// M2.7 released Mar 18, 2026 - Self-evolving, #1 AI Intelligence Index, SWE-Pro 56.22%
+// M3 released May 31, 2026 - 1M ctx, MSA sparse attention (~1/20 compute at 1M), native multimodal
 export const MINIMAX_MODELS = {
-    M2_7: "minimax/minimax-m2.7", // 2300B/100B MoE, 200K ctx, SWE-Pro 56.22%, Multi-SWE #1, $0.30/$1.20
+    M3: "minimax/minimax-m3", // 1M ctx, MSA sparse attention, multimodal, agentic/coding, $0.30/$1.20
+    M2_7: "minimax/minimax-m2.7", // 2300B/100B MoE, 200K ctx, SWE-Pro 56.22%, Multi-SWE #1 (fallback)
     M2_5: "minimax/minimax-m2.5", // SWE-Bench 80.2%, 37% faster than M2.1 (legacy)
     M2_1: "minimax/minimax-m2.1", // 230B/10B MoE - SWE-Bench 72.5% (legacy)
 };
@@ -185,10 +188,10 @@ export const CURRENT_MODELS = {
         reason: PERPLEXITY_MODELS.SONAR_REASONING, // sonar-reasoning-pro $2/$8 per M
     },
     openrouter: {
-        kimi: KIMI_MODELS.K2_6, // K2.6 (Apr 2026): 1T MoE, SWE-bench Pro leader
+        kimi: KIMI_MODELS.K2_7_CODE, // K2.7-Code (Jun 2026): coding-specialized, built on K2.6
         qwen: QWEN_MODELS.CODER_NEXT, // Qwen3-Coder-Next: 80B/3B MoE, 262K ctx, SWE >70% (no 3.6-coder yet)
         qwen_reason: QWEN_MODELS.MAX_THINKING, // 235B MoE thinking mode (HMMT 98%) — still best for reasoning
-        minimax: MINIMAX_MODELS.M2_7, // M2.7: SWE-Pro 56.22%, Multi-SWE #1, self-evolving
+        minimax: MINIMAX_MODELS.M3, // M3: 1M ctx, MSA sparse attention, agentic/coding
     }
 };
 // Tool-specific defaults - References CURRENT_MODELS for easy bumping
@@ -303,18 +306,18 @@ export const TOOL_DEFAULTS = {
         temperature: 0.3, // Lower for precise reasoning
     },
     kimi_thinking: {
-        model: KIMI_MODELS.K2_6, // K2.6 (Apr 2026): 1T MoE, SWE-bench Pro leader
+        model: KIMI_MODELS.K2_7_CODE, // K2.7-Code (Jun 2026): coding-specialized, built on K2.6
         maxTokens: 16000,
         temperature: 0.7,
     },
     // MiniMax tools - VERY CHEAP, open source
     minimax_code: {
-        model: MINIMAX_MODELS.M2_7, // SWE-Pro 56.22%, Multi-SWE #1
+        model: MINIMAX_MODELS.M3, // 1M ctx, MSA sparse attention, agentic/coding
         maxTokens: 4000,
         temperature: 0.3, // Lower for precise code
     },
     minimax_agent: {
-        model: MINIMAX_MODELS.M2_7, // SWE-Pro 56.22%, self-evolving, #1 AI Intelligence Index
+        model: MINIMAX_MODELS.M3, // 1M ctx, multimodal, long-horizon agentic work
         maxTokens: 4000,
         temperature: 0.5, // Balanced for agentic tasks
     },
@@ -369,6 +372,7 @@ export const MODEL_DISPLAY_NAMES = {
     "sonar-reasoning-pro": "perplexity-reason",
     // Kimi (Moonshot)
     "moonshotai/kimi-k2-thinking": "kimi-k2",
+    "moonshotai/kimi-k2.7-code": "kimi-k2.7-code",
     "moonshotai/kimi-k2.6": "kimi-k2.6",
     "moonshotai/kimi-k2.5": "kimi-k2.5",
     "moonshotai/kimi-k2.5-thinking": "kimi-k2.5",
@@ -381,6 +385,7 @@ export const MODEL_DISPLAY_NAMES = {
     "qwen/qwen3-max-thinking": "qwen-max",
     "qwen/qwen3-235b-a22b-thinking-2507": "qwen-235b-thinking",
     // MiniMax
+    "minimax/minimax-m3": "minimax-m3",
     "minimax/minimax-m2.7": "minimax-m2.7",
     "minimax/minimax-m2.5": "minimax-m2.5",
     "minimax/minimax-m2.1": "minimax-m2.1",
@@ -422,6 +427,7 @@ export const MODEL_PRICING = {
     "sonar-reasoning-pro": 0.006, // avoid - expensive reasoning tokens
     // OpenRouter models - Kimi
     "moonshotai/kimi-k2-thinking": 0.002,
+    "moonshotai/kimi-k2.7-code": 0.002125, // ($0.75 + $3.50) / 2 / 1000 (Jun 12, 2026)
     "moonshotai/kimi-k2.6": 0.0027, // ($0.74 + $4.65) / 2 / 1000 (Apr 2026)
     "moonshotai/kimi-k2.5": 0.003,
     "moonshotai/kimi-k2.5-thinking": 0.003,
@@ -434,7 +440,8 @@ export const MODEL_PRICING = {
     "qwen/qwen3-max-thinking": 0.005,
     "qwen/qwen3-235b-a22b-thinking-2507": 0.000822, // ($0.15 + $1.50) / 2 / 1000
     // OpenRouter models - MiniMax (VERY CHEAP!)
-    "minimax/minimax-m2.7": 0.00075, // ($0.30 + $1.20) / 2 / 1000 - flagship
+    "minimax/minimax-m3": 0.00075, // ($0.30 + $1.20) / 2 / 1000 - flagship (promo pricing)
+    "minimax/minimax-m2.7": 0.00075, // ($0.30 + $1.20) / 2 / 1000 - fallback
     "minimax/minimax-m2.5": 0.000685, // legacy
     "minimax/minimax-m2.1": 0.000685, // legacy
 };

package/dist/src/config/model-defaults.js

@@ -28,8 +28,8 @@ const MODELS = {
     // OpenRouter
     QWEN: QWEN_MODELS.CODER_NEXT, // qwen/qwen3-coder-next (80B/3B MoE, 262K ctx)
     QWEN_REASON: QWEN_MODELS.MAX_THINKING, // qwen/qwen3-235b-a22b-thinking (235B MoE)
-    KIMI: KIMI_MODELS.K2_6, // moonshotai/kimi-k2.6 (Apr 2026 - SWE-bench Pro leader)
-    MINIMAX: MINIMAX_MODELS.M2_7, // minimax/minimax-m2.7 (SWE-Pro 56.22%, #1 AI Intelligence Index)
+    KIMI: KIMI_MODELS.K2_7_CODE, // moonshotai/kimi-k2.7-code (Jun 2026 - coding-specialized, built on K2.6)
+    MINIMAX: MINIMAX_MODELS.M3, // minimax/minimax-m3 (May 2026 - 1M ctx, MSA sparse attention)
 };
 /**
  * Get Scout model configuration

package/dist/src/config/timeout-config.js

@@ -36,8 +36,8 @@ export function getTimeoutConfig() {
  * Thinking/reasoning/swarm models get extended timeout (600s default).
  * Standard models get 180s default.
  *
- * NOTE: All Kimi K2 variants (incl. 'moonshotai/kimi-k2.6') run the Agent Swarm
- * and need the extended timeout — but their model IDs don't contain 'thinking',
+ * NOTE: All Kimi K2 variants (incl. 'moonshotai/kimi-k2.7-code', which is always in
+ * thinking mode) need the extended timeout — but their model IDs don't contain 'thinking',
  * so we match 'kimi' explicitly. Same applies to DeepSeek V4 and Zhipu GLM, whose
  * reasoning passes are slow but whose IDs lack the 'thinking'/'reasoning' marker.
  * This keeps primary and fallback consistent and lets these tools inherit 600s

package/dist/src/memory/index.js

@@ -9,7 +9,7 @@ export * from './memory-manager.js';
 // Provider exports
 export * from './providers/local-provider.js';
 export * from './providers/mem0-provider.js';
-// export * from './providers/dokoro-provider.js';
+export * from './providers/dokoro-provider.js';
 // export * from './providers/hybrid-provider.js';
 // Main API
 export { getMemoryManager, resetMemoryManager, HierarchicalMemoryManager } from './memory-manager.js';

package/dist/src/memory/memory-config.js

@@ -109,9 +109,8 @@ export function validateMemoryConfig(config) {
     if (config.provider === 'mem0' && !config.mem0?.apiKey && !process.env.MEM0_API_KEY) {
         errors.push('Mem0 provider requires API key (MEM0_API_KEY or config.mem0.apiKey)');
     }
-    if (config.provider === 'dokoro' && !config.dokoro?.connectionString && !process.env.DOKORO_CONNECTION) {
-        errors.push('Dokoro provider requires connection string');
-    }
+    // Dokoro is file-backed: connectionString is optional (falls back to
+    // DOKORO_PATH env or {cwd}/dokoro), so no validation needed here.
     if (config.provider === 'local' && !config.local?.path) {
         errors.push('Local provider requires storage path');
     }

package/dist/src/memory/memory-manager.js

@@ -6,7 +6,7 @@ import { mergeMemoryConfig, validateMemoryConfig } from './memory-config.js';
 import { LocalProvider } from './providers/local-provider.js';
 import { Mem0Provider } from './providers/mem0-provider.js';
 import { randomBytes } from 'crypto';
-// import { DokoroProvider } from './providers/dokoro-provider.js';
+import { DokoroProvider } from './providers/dokoro-provider.js';
 // import { HybridProvider } from './providers/hybrid-provider.js';
 /**
  * Main memory manager that coordinates all providers
@@ -245,9 +245,15 @@ export class HierarchicalMemoryManager {
                     return null;
                 }
             case 'dokoro':
-                // TODO: Implement Dokoro provider
-                console.warn('Dokoro provider not yet implemented');
-                return null;
+                try {
+                    const provider = new DokoroProvider(this.config.dokoro || {});
+                    await provider.initialize();
+                    return provider;
+                }
+                catch (error) {
+                    console.error('Failed to create Dokoro provider:', error);
+                    return null;
+                }
             case 'local':
                 return await this.createLocalProvider();
             case 'hybrid':