[PR #337] Add Kimi Code integration guide for DeepSeek #353

Open
opened 2026-10-26 09:48:21 +00:00 by bsakaguchi · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/awesome-deepseek-agent/pull/337
Author: @CislinkNL
Created: 8/1/2026
Status: 🔄 Open

Base: main ← Head: feat/kimi-code-guide


📝 Commits (2)

  • f148767 Add Kimi Code integration guide for DeepSeek
  • 9bb4180 Address review: unify reasoning effort levels, clarify env sub-table

📊 Changes

4 files changed (+236 additions, -0 deletions)

View changed files

📝 README.md (+1 -0)
📝 README.zh-CN.md (+1 -0)
➕ docs/kimi_code.md (+117 -0)
➕ docs/kimi_code.zh-CN.md (+117 -0)

📄 Description

What

Adds a bilingual integration guide for Kimi Code — Moonshot AI's terminal AI coding agent — covering installation → DeepSeek provider configuration → first run.

  • docs/kimi_code.md (English)
  • docs/kimi_code.zh-CN.md (Simplified Chinese)
  • README table entries in both README.md and README.zh-CN.md (inserted in alphabetical order between Kilo Code and Langcli)

What the guide covers

  • Install: official install script (macOS/Linux/Windows PowerShell) and npm (@moonshot-ai/kimi-code, Node 22.19+)
  • Configure: DeepSeek provider via the OpenAI-compatible protocol (type = "openai", base_url = "https://api.deepseek.com/v1"), with a tip to keep the API key in the provider's env sub-table instead of plain text
  • Models: deepseek-v4-pro / deepseek-v4-flash (current V4 names), each declared with max_context_size = 1000000 to reflect the 1M context window
  • Thinking: [thinking] section with effort = "max", noting Kimi Code auto-handles reasoning_content / reasoning_effort for OpenAI-compatible providers
  • Usage: TUI (kimi), single-shot (kimi -p), /model switching, /provider management

Checklist

  • Bilingual: EN + zh-CN in a single PR
  • README table entries in both files, alphabetical order
  • Uses current model names (deepseek-v4-pro / deepseek-v4-flash), no deprecated deepseek-chat
  • Reflects 1M context window (max_context_size = 1000000)
  • Compatible with max thinking effort
  • Configuration verified against the official Kimi Code docs (providers, config files) and by real-world usage

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/deepseek-ai/awesome-deepseek-agent/pull/337 **Author:** [@CislinkNL](https://github.com/CislinkNL) **Created:** 8/1/2026 **Status:** 🔄 Open **Base:** `main` ← **Head:** `feat/kimi-code-guide` --- ### 📝 Commits (2) - [`f148767`](https://github.com/deepseek-ai/awesome-deepseek-agent/commit/f148767b2459dd7d061a415a5e96c1b7a241c952) Add Kimi Code integration guide for DeepSeek - [`9bb4180`](https://github.com/deepseek-ai/awesome-deepseek-agent/commit/9bb41808a9049b81ec549c3d5ba8b06a6408f9e4) Address review: unify reasoning effort levels, clarify env sub-table ### 📊 Changes **4 files changed** (+236 additions, -0 deletions) <details> <summary>View changed files</summary> 📝 `README.md` (+1 -0) 📝 `README.zh-CN.md` (+1 -0) ➕ `docs/kimi_code.md` (+117 -0) ➕ `docs/kimi_code.zh-CN.md` (+117 -0) </details> ### 📄 Description ## What Adds a bilingual integration guide for [Kimi Code](https://www.kimi.com/code/docs/en/) — Moonshot AI's terminal AI coding agent — covering installation → DeepSeek provider configuration → first run. - `docs/kimi_code.md` (English) - `docs/kimi_code.zh-CN.md` (Simplified Chinese) - README table entries in both `README.md` and `README.zh-CN.md` (inserted in alphabetical order between Kilo Code and Langcli) ## What the guide covers - **Install**: official install script (macOS/Linux/Windows PowerShell) and npm (`@moonshot-ai/kimi-code`, Node 22.19+) - **Configure**: DeepSeek provider via the OpenAI-compatible protocol (`type = "openai"`, `base_url = "https://api.deepseek.com/v1"`), with a tip to keep the API key in the provider's `env` sub-table instead of plain text - **Models**: `deepseek-v4-pro` / `deepseek-v4-flash` (current V4 names), each declared with `max_context_size = 1000000` to reflect the 1M context window - **Thinking**: `[thinking]` section with `effort = "max"`, noting Kimi Code auto-handles `reasoning_content` / `reasoning_effort` for OpenAI-compatible providers - **Usage**: TUI (`kimi`), single-shot (`kimi -p`), `/model` switching, `/provider` management ## Checklist - [x] Bilingual: EN + zh-CN in a single PR - [x] README table entries in both files, alphabetical order - [x] Uses current model names (`deepseek-v4-pro` / `deepseek-v4-flash`), no deprecated `deepseek-chat` - [x] Reflects 1M context window (`max_context_size = 1000000`) - [x] Compatible with `max` thinking effort - [x] Configuration verified against the official Kimi Code docs ([providers](https://www.kimi.com/code/docs/en/kimi-code-cli/configuration/providers.html), [config files](https://www.kimi.com/code/docs/en/kimi-code-cli/configuration/config-files.html)) and by real-world usage --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
Sign in to join this conversation.
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/awesome-deepseek-agent#353
No description provided.