Search Results for "kimi" | AgentCC

nowait-reasoning-optimizer

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Use when optimizing inference of reasoning models (QwQ, DeepSeek-R1, Phi4-Reasoning, Qwen3, Kimi-VL, QvQ), reducing chain-of-thought token usage by 27-51% while preserving accuracy. Triggers on "optimize reasoning", "reduce thinking tokens", "efficient inference", "suppress reflection tokens", or when working with verbose CoT outputs.

gen-changelog

Generate changelog entries for code changes.

pull-request

Create and submit a GitHub Pull Request.

release

Execute the release workflow for Kimi Code CLI packages.

gen-docs

Update Kimi Code CLI user documentation.

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Kimi's capabilities with specialized knowledge, workflows, or tool integrations.

translate-docs

Translate and sync bilingual documentation.

kimi-cli-help

Answer Kimi Code CLI usage, configuration, and troubleshooting questions. Use when user asks about Kimi Code CLI installation, setup, configuration, slash commands, keyboard shortcuts, MCP integration, providers, environment variables, how something works internally, or any questions about Kimi Code CLI itself.

kimi-integration

Step-by-step guide for integrating Moonshot AI (Kimi) and Kimi Code models into Clawdbot. Use when someone asks how to add Kimi models, configure Moonshot AI, or set up Kimi for Coding in Clawdbot.

ccs-delegation

Auto-activate CCS CLI delegation for deterministic tasks. Parses user input, auto-selects optimal profile (glm/kimi/custom) from ~/.ccs/config.json, enhances prompts with context, executes via `ccs {profile} -p "task"` or `ccs {profile}:continue`, and reports results. Triggers on "use ccs [task]" patterns, typo/test/refactor keywords. Excludes complex architecture, security-critical code, performance optimization, breaking changes.

proxy-mode-reference

Reference guide for using external AI models via claudish CLI. Use when running multi-model reviews, understanding how /team invokes external models, or debugging external model integration issues. Includes routing prefixes for MiniMax, Kimi, GLM direct APIs.

multi-model-validation

Run multiple AI models in parallel for 3-5x speedup with ENFORCED performance statistics tracking. Use when validating with Grok, Gemini, GPT-5, DeepSeek, MiniMax, Kimi, GLM, or Claudish proxy for code review, consensus analysis, or multi-expert validation. NEW in v3.2.0 - Direct API prefixes (mmax/, kimi/, glm/) for cost savings. Includes dynamic model discovery via `claudish --top-models` and `claudish --free`, session-based workspaces, and Pattern 7-8 for tracking model performance. Trigger keywords - "grok", "gemini", "gpt-5", "deepseek", "minimax", "kimi", "glm", "claudish", "multiple models", "parallel review", "external AI", "consensus", "multi-model", "model performance", "statistics", "free models".

orchestrationclaudishparallel+11

llm-models

Access Claude, Gemini, Kimi, GLM and 100+ LLMs via inference.sh CLI using OpenRouter. Models: Claude Opus 4.5, Claude Sonnet 4.5, Claude Haiku 4.5, Gemini 3 Pro, Kimi K2, GLM-4.6, Intellect 3. One API for all models with automatic fallback and cost optimization. Use for: AI assistants, code generation, reasoning, agents, chat, content generation. Triggers: claude api, openrouter, llm api, claude sonnet, claude opus, gemini api, kimi, language model, gpt alternative, anthropic api, ai model api, llm access, chat api, claude alternative, openai alternative

kb-kimi

Kimi K2（Moonshot AI）特有の問題とワークアラウンド。Claudeモデルとの差異、リトライ処理等

qa_review_triple-model

Launch four independent AI code reviewers (Opus, Gemini, Codex, Kimi K2.5) to QA/QC code or notebooks. Each reviewer writes findings to separate markdown files, then orchestrator synthesizes. Use for critical code review, bug investigation, or quality assurance tasks. Triggers: triple review, quad review, four model review, independent code review, QAQC, quality assurance, multi-model analysis, cross-validation, bug investigation, critical review, kimi review, togetherai review

dev_invoke_kimi-cli

Delegate testing, QA, and code review tasks to Opencode CLI using Kimi K2.5 model via markdown file handoff. Write test request to TASK.md, Opencode with Kimi K2.5 generates tests/reviews, outputs to OUTPUT.md. Use for test generation, QA verification, edge case detection, code coverage analysis, security reviews. Triggers: kimi, kimi k2.5, kimi cli, opencode kimi, test generation, QA, quality assurance, code review, unit tests, integration tests, edge cases, test coverage, testing, kimisubagent, togetherai kimi, opencode/kimi-k2.5-free, togetherai/moonshotai/Kimi-K2.5 Prerequisites: Opencode CLI installed, Together.ai API key (if using togetherai provider) Models: opencode/kimi-k2.5-free (recommended), togetherai/moonshotai/Kimi-K2.5 (alternative)

moonshot-ai

Moonshot AI Kimi API - Trillion-parameter MoE model with 256K context, tool calling, and agentic capabilities for chat, coding, and autonomous task execution

ask

Answer questions about Pull Request code changes.

code-review

AI-powered code review focusing on bugs, security, and performance

agent-setup

gastown-publish

Set up Claude Code, Kimi Code, and OpenClaw sub-agent instances. Use when configuring new agent accounts, installing plugins across accounts, or setting up permissions for autonomous operation.

ai-article

自动搜集AI领域热点或根据指定选题，按照二哥的写作风格完成AI技术类文章撰写。专注于AI Coding工具实测（Claude Code、Qoder、Cursor、TRAE等）、大模型工程化落地（SpringAI、LangChain、RAG等）、AI Agent和工作流编排、国产大模型评测（GLM、通义千问、DeepSeek、MiniMax、Kimi等）、各种AI工具、Agent工具的评测。触发关键词：写一篇AI文章、AI技术文章、大模型测评、AI工具实测、GLM、Claude Code、Qoder、Cursor、TRAE、SpringAI、RAG、Agent、工作流、国产大模型、搜集AI热点、AI选题等。

ais-orchestrator

gastown-publish

Orchestrate many Claude Code and Kimi Code sub-agents in parallel via tmux. Use when: decomposing large tasks into parallel work, managing multiple coding agents, monitoring agent health, rotating accounts on rate limits. Requires: ais CLI (~/.local/bin/ais), tmux.

Together Chat Completions

Serverless chat and text completion inference via Together AI's OpenAI-compatible API. Access 100+ open-source models with pay-per-token pricing. Includes function calling (tool use), structured outputs (JSON mode, json_schema, regex), and reasoning models (DeepSeek R1, Qwen3 Thinking, Kimi K2). Use when building chat applications, text generation, multi-turn conversations, function calling, structured JSON outputs, reasoning/chain-of-thought, or any LLM inference task using Together AI.

china-model-selection-guide

China model selection and task-routing guide for Doubao-Seed-2.0-Code, GLM-5, MiniMax-M2.5, and Kimi-K2.5. Use when users need to choose the best-fit model by input type, task complexity, engineering constraints, and delivery goals, including staged multi-model workflows.

video-analyzer

Analyze video content using Moonshot Kimi-k2.5 multimodal model and generate structured reports.

OpenCode

Use OpenCode as the primary coding agent. OpenCode is an open-source AI coding agent with specialized agent modes (plan/build), a subagent system, and headless server capabilities. Use cases: (1) planning and analysis, (2) building and implementing features, (3) code reviews, (4) long-running background tasks, (5) API-driven automation. Supports Kimi, Claude, GPT, Gemini and 75+ providers. Interactive mode requires the bash tool with pty:true.

gastown-knowledge

gastown-publish

Gas Town system knowledge — architecture, accounts, tools, and workflows. Use when an agent asks about Gas Town infrastructure, Claude/Kimi accounts, gt commands, tmux agent patterns, or operational setup.

Together Reasoning

Use reasoning and thinking models on Together AI including DeepSeek R1, DeepSeek V3.1, Kimi K2-Thinking, and Qwen3 with thinking mode. Models output chain-of-thought in <think>...</think> tags before answering. Adjustable reasoning effort (low/medium/high). Use when users need reasoning models, chain-of-thought, step-by-step thinking, math/code/logic problem solving, or adjustable reasoning depth.

knowpatch

LLM knowledge cutoff compensator — knowledge corrections for breaking changes and API drift. Covers: renamed packages (shadcn-ui→shadcn), changed APIs (z.string().email()→z.email()), new config formats (.eslintrc→eslint.config.js), current model IDs, Apple platform changes. Versions are NOT cached — always verify via package manager. Useful for: install, create, scaffold, upgrade, migrate, latest, version, dependency, shadcn, tailwind, vite, zod, eslint, react, next, svelte, nuxt, django, fastapi, pydantic, ruff, uv, claude, gpt, gemini, openai, anthropic, deepseek, qwen, llama, mistral, kimi, minimax, glm, node, bun, python, typescript, macos, xcode, swift, swiftui, supabase, npm, pip, cargo, package.json, pyproject.toml, requirements.txt.