v0.18.0 — reliability & performance pass

Your codebase,
fully understood.

Chat, search, explain, refactor, audit, and generate board-ready reports — with source citations, confidence scores, and severity labels on every claim. Local-first. French & English.

Get started View on GitHub

Works everywhere you code

DeepSeek Harness Cursor Claude Windsurf Any MCP client Terminal CLI

dsh-codebase-chat — DeepSeek Harness

EXPLORER

my-app

src

App.tsx

utils.ts

api.ts

package.json

README.md

src/App.tsx

1import React from 'react';

3const TOKEN = "abc123"; // ⚠️ hardcoded

5export default function App() {

6 return <main className="app" />;

Codebase Pro

dsh-codebase-chat

▋

- const TOKEN = "abc123";
+ const TOKEN = process.env.TOKEN;

Severity: High Confidence: 94% src/App.tsx:3

dsh-codebase-chat

●

MCP tools exposed

Packages — plugin & MCP

100%

Local-first indexing

FR / EN

Bilingual out of the box

How it works

From messy code to clear decisions

A four-stage pipeline runs entirely on your machine before a single token reaches the model.

Index

AST + regex extraction builds a symbol-level index of your repo — incrementally, cached on disk.

Retrieve

Lexical + optional local semantic embeddings pick the most relevant chunks inside a token budget.

Prompt

A sourced prompt is assembled with module graphs, metrics, and citations — in FR or EN.

Decide

Every claim comes back with [source: file:line], confidence, and severity.

Capabilities

Built for professional code reviews

Not just a chat box — a full intelligence layer over your repository.

Chat & Search

Ask natural questions, find symbols, explain files — every answer cites file and line.

Intelligence Pro

CTO-level brief: architecture, tech radar, debt map, security posture, competitor signals.

Audit & Tasks

Detect TODO/FIXME/any/debugger and generate a sprinted TASKS.md with Before/After patches.

Refactor & Apply

Production-ready refactors applied with backup, dry-run, verification, and rollback.

Strategic Reports

Board-ready SWOT, score cards, 90-day roadmap, CEO brief, and marketing positioning.

MCP Compatible

Standalone server for Cursor, Claude, Windsurf — with a key-free prompt-only mode.

Distribution

Two packages, one engine

Same intelligence core — pick the integration that fits your workflow.

dsh-codebase-chat

DeepSeek Harness plugin + CLI

✓Codebase Pro panel inside DeepSeek Harness
✓Slash commands: /codebase-intel, /codebase-audit…
✓Standalone CLI for any project
✓Safe apply: backup, dry-run, rollback

npm install dsh-codebase-chat

dsh-codebase-chat-mcp

Standalone MCP server

✓11 tools over stdio — works in Cursor, Claude, Windsurf
✓Prompt-only mode — no API key needed, the host model answers
✓Optional direct LLM calls via DeepSeek / OpenAI-compatible API
✓Local multilingual embeddings for semantic retrieval

npx dsh-codebase-chat-mcp

Reference

The full tool suite

Every tool is available in DeepSeek Harness and through the MCP server.

Tool	What it does
codebase_chat	Q&A on your project with cited sources
codebase_search	Find symbols, terms, and usages across the repo
codebase_explain	Explain a file or a symbol — fallback symbol search included
codebase_refactor	Propose a concrete refactor with Before/After diff
codebase_intelligence	Full CTO brief: architecture, debt, security, roadmap
codebase_audit	Tech-debt and non-conformity scan with severity
codebase_report	Strategic board report — SWOT, risks, opportunities
codebase_ceo	One-page executive brief
codebase_tasks	Generate a sprinted TASKS.md action plan
codebase_player	UX / playthrough brief from a user's point of view
codebase_crea	Creative and marketing angles extracted from the code

All tools accept projectPath, lang (fr/en), embed, and promptOnly.

See it in action

CLI demo

Index a project and ask questions from your terminal.

Get started

Install

Three ways to run it. All read code locally.

1DeepSeek Harness plugin

pnpm add dsh-codebase-chat
# or
npm install dsh-codebase-chat

2Standalone MCP server

npm install -g dsh-codebase-chat-mcp
npx dsh-codebase-chat-mcp

3From source

git clone https://github.com/shinzarou-eng/dsh-codebase-chat.git
cd dsh-codebase-chat
pnpm install

Configuration

MCP configuration

Add this block to your IDE's MCP server settings. No API key needed — the host model answers.

mcp.json

{
  "mcpServers": {
    "dsh-codebase-chat": {
      "command": "npx",
      "args": ["dsh-codebase-chat-mcp"]
    }
  }
}

Optional env: DEEPSEEK_API_KEY, OPENAI_API_KEY, CODEBASE_MODEL — only for direct LLM calls.

Slash commands

One command per job

Run them headless in CI or interactively inside DeepSeek Harness.

terminal

$ dsh --profile headless '/codebase-intel'  --project C:\my-app
$ dsh --profile headless '/codebase-audit'  --project C:\my-app --lang en
$ dsh --profile headless '/codebase-report' --project C:\my-app
$ dsh --profile headless '/codebase-tasks'  --project C:\my-app --apply
$ dsh --profile headless '/codebase-ceo'    --project C:\my-app
$ dsh --profile headless '/codebase-player' --project C:\my-app

Trust

Evidence & scoring

Every technical claim, risk, and opportunity is backed by a source citation and scored for confidence and severity.

[source: src/App.tsx:42]

[Confidence: 95%]

[Severity: High]

What's next

Roadmap

Organized by horizon, not by date. Full detail in ROADMAP.md.

NEXT Engine depth

→Tree-sitter parsing — Python, Go, Rust, Java, C#
→Watch mode — hot index while you code
→GitHub Issues export from TASKS.md
→Diff-aware retrieval on changed files
→Per-project config file
→Prompt packs — ES, DE, PT

PLANNED Integrations

→VS Code extension — sidebar + inline explain
→HTTP/SSE transport for the MCP server
→PR review mode via GitHub API
→Report export — standalone HTML / PDF
→Interactive diff viewer in Codebase Pro

EXPLORING Longer horizon

→Multi-repo workspaces
→Shared team index cache
→CI bot — audit summaries on every PR
→Local reranker for tighter context
→JetBrains & Neovim clients

Questions

FAQ

Does it send my code to the cloud?

No. Indexing, retrieval, and prompt building all run on your machine. In prompt-only MCP mode, nothing leaves your machine at all — the host model reads the assembled context directly.

Do I need an API key?

Not for the MCP server: without a key it returns the built prompt to the host model (promptOnly). A DEEPSEEK_API_KEY or OPENAI_API_KEY is only needed if you want the server to call the LLM itself. Inside DeepSeek Harness, the plugin uses your configured model.

Can I use it with Claude, Cursor, or Windsurf?

Yes. Install dsh-codebase-chat-mcp and add it to your MCP server config — see the configuration block above.

Does it work without DeepSeek Harness?

Yes. The dsh-codebase-chat-mcp package is fully standalone, and the main package ships a CLI that works on any project.

Which languages are supported?

French and English out of the box via the lang parameter on every tool and command. The prompt builders are designed to be extended.

Is applying patches safe?

Yes. Apply workflows support dry-run, create .dsh-backups/ copies before overwriting, enforce protected paths, and keep all writes inside the selected project.

Ready to understand your codebase?

One install. Eleven tools. Zero cloud upload.

Get started Star on GitHub

Your codebase, fully understood.

From messy code to clear decisions

Index

Retrieve

Prompt

Decide

Built for professional code reviews

Chat & Search

Intelligence Pro

Audit & Tasks

Refactor & Apply

Strategic Reports

MCP Compatible

Two packages, one engine

dsh-codebase-chat

dsh-codebase-chat-mcp

The full tool suite

CLI demo

Install

1DeepSeek Harness plugin

2Standalone MCP server

3From source

MCP configuration

One command per job

Evidence & scoring

Roadmap

FAQ

Ready to understand your codebase?

Your codebase,
fully understood.