Reasoning Traces

Give your coding agent a stronger brain to consult.

Reasoning Traces is an MCP server + Claude Code plugin. It adds a deep_reasoning tool: the agent sends a hard problem (plus the code and context it has gathered) to a stronger reasoning model, gets back the model's full reasoning trace, and uses that trace to shape and cross-check its own answer.

Works out of the box with reasoning models on OpenRouter (default: DeepSeek R1, which returns its complete raw chain of thought). Anthropic and custom backends included.

Install (Claude Code)

Prerequisites: uv (curl -LsSf https://astral.sh/uv/install.sh | sh) and an OpenRouter API key.

1. Export your key (add to ~/.zshrc / ~/.bashrc to persist):

   export OPENROUTER_API_KEY=sk-or-v1-...

2. In Claude Code:

   /plugin marketplace add dhruv-corethink/reasoning-traces
   /plugin install reasoning-traces@corethink

3. Restart Claude Code (or start a new session). Done — the plugin works in every project.

Verify with /mcp (the reasoning-traces server should be connected).

Related MCP server: Deepseek Thinker MCP Server

Usage

• Automatic — Claude Code calls deep_reasoning on its own when a task involves multi-step reasoning (subtle bugs, architecture trade-offs, algorithm design, math). The tool description steers this.

• On demand — force a consultation:

  /reason why does this async queue deadlock under load?

The tool result contains the reasoning model's full trace plus its conclusion; Claude Code verifies it against your actual code before answering.

Team rollout (zero-command install)

Add this to a shared repo's .claude/settings.json and every teammate gets the plugin automatically when they trust the workspace:

{
  "extraKnownMarketplaces": {
    "corethink": {
      "source": { "source": "github", "repo": "dhruv-corethink/reasoning-traces" }
    }
  },
  "enabledPlugins": { "reasoning-traces@corethink": true }
}

Each teammate still needs their own OPENROUTER_API_KEY in their environment.

Configuration

Set env vars in your shell, or per-project in a .env file (the server loads .env from the working directory; existing env vars win).

DeepSeek R1 is the default because it returns its full raw reasoning trace; most other models (o3, Gemini) return summaries.

Other MCP clients

Any MCP client (Claude Desktop, Cursor, etc.) can run the server without the plugin:

{
  "mcpServers": {
    "reasoning-traces": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/dhruv-corethink/reasoning-traces", "reasoning-traces"],
      "env": { "OPENROUTER_API_KEY": "sk-or-v1-..." }
    }
  }
}

Or with plain Claude Code CLI, no plugin:

claude mcp add --scope user reasoning-traces -- uvx --from git+https://github.com/dhruv-corethink/reasoning-traces reasoning-traces

Custom backends

reasoning_traces/backends.py defines a tiny interface — reason(prompt) -> ReasoningResult(trace, conclusion). Three backends ship today:

• openrouter (default) — any reasoning model on OpenRouter

• anthropic — Claude Opus 4.8 with adaptive thinking (summarized reasoning; the Anthropic API never exposes raw chain of thought)

• corethink — stub for the Corethink reasoning model (coming soon)

Development

git clone https://github.com/dhruv-corethink/reasoning-traces
cd reasoning-traces
echo "OPENROUTER_API_KEY=sk-or-v1-..." > .env   # gitignored

Open Claude Code in the repo — .mcp.json runs the server straight from source via uvx. The .env is loaded by the server at startup.

License

MIT