opus-advisor-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@opus-advisor-mcp review this database migration for potential race conditions"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
opus-advisor-mcp
An MCP server that lets Claude Code consult Opus as a strategic advisor mid-task. Run your session on Sonnet or Haiku, and escalate complex decisions to Claude Opus 4.7 on demand — using your existing Claude Code subscription.
Inspired by Anthropic's Advisor Strategy.
How it works
┌─────────────────────────────────────────────┐
│ Claude Code (Sonnet) │
│ │
│ "I need to decide on the DB schema..." │
│ │ │
│ ▼ │
│ calls consult_opus MCP tool │
│ │ │
└────────┼────────────────────────────────────┘
│
▼
┌─────────────────────────────────────────────┐
│ opus-advisor MCP server │
│ │
│ 1. Reads prior consultation history │
│ 2. Reads requested files from disk │
│ 3. Pipes prompt to: claude -p --model opus │
│ 4. Logs advice to advisor-log.md │
│ 5. Returns advice to Sonnet │
└─────────────────────────────────────────────┘No API keys needed. The server shells out to the claude CLI, which uses your existing authentication.
Related MCP server: codex-bridge
Install
npm install -g opus-advisor-mcpOr clone and build locally:
git clone https://github.com/Divinci-AI/opus-advisor-mcp.git
cd opus-advisor-mcp
npm install
npm run buildConfigure
Add to your project's .mcp.json or ~/.claude/.mcp.json:
{
"mcpServers": {
"opus-advisor": {
"command": "opus-advisor",
"timeout": 180000
}
}
}If installed locally (not globally):
{
"mcpServers": {
"opus-advisor": {
"command": "node",
"args": ["/path/to/opus-advisor-mcp/dist/index.js"],
"timeout": 180000
}
}
}Restart Claude Code after adding the config.
Tools
consult_opus
Consult Opus for strategic advice.
Parameter | Type | Default | Description |
| string | required | The question or problem you need advice on |
| string | optional | Additional context, constraints, or background |
| string[] | optional | File paths (relative to project root) to include as code context |
|
|
| Reasoning effort level for Opus |
| boolean |
| Include prior consultation history for continuity |
Example:
{
"question": "Is this database migration safe under concurrent writes?",
"files": ["src/db/migration-042.ts", "src/db/schema.ts"],
"effort": "high"
}read_advisor_log
Read the consultation log from prior calls.
Parameter | Type | Description |
| number | Number of recent consultations to return (omit for all) |
read_advisor_meta
Read structured metadata (latency, token counts, effort levels).
Parameter | Type | Description |
| number | Number of recent entries to return (omit for all) |
clear_advisor_log
Clear the consultation log and metadata to start fresh.
Features
No API key required — Uses your existing Claude Code subscription via the
claudeCLIPer-project logs — Consultation history is stored per project at
~/.opus-advisor/<project>-<hash>/Code-aware context — Pass file paths directly; the server reads and injects them as labeled code blocks
Consultation continuity — Prior advice is fed back as context so Opus can build on earlier decisions
Token-aware history — History is capped by both entry count (5) and token budget (~6K tokens)
Metadata tracking — Latency, token estimates, and effort levels tracked in
advisor-meta.jsonlSignal protection — Partial output from killed processes is discarded, not returned as advice
Path traversal guard — File reads are validated to stay within the project root
Security
Path traversal protection: The
filesparameter validates that all resolved paths remain within the project root directory. Paths like../../etc/passwdor absolute paths outside the project are rejected.Binary file filtering: Common binary extensions (images, executables, archives, etc.) are automatically skipped.
No shell execution: The server uses
spawnwith array arguments and pipes the prompt via stdin. No shell interpolation occurs.Local only: The MCP server runs locally via stdio. No network ports are opened.
Consultation logs: Stored at
~/.opus-advisor/in plaintext. These may contain code snippets and questions from your consultations. Do not commit or share these files if they contain sensitive code.
Environment Variables
Variable | Description |
| Override the log directory (default: |
How it compares to Anthropic's Advisor Tool
Anthropic's advisor_20260301 is a server-side API feature where the advisor sees the full conversation transcript within a single API request. This MCP server is a different approach:
Anthropic Advisor Tool | opus-advisor-mcp | |
Context sharing | Full transcript (server-side) | Question + files + history (client-side) |
Auth | API key required | Uses existing Claude Code subscription |
Integration | API-level ( | MCP tool (works in Claude Code today) |
Persistence | None | Markdown log + JSONL metadata |
Cost | Billed per-token at Opus rates | Included in subscription |
Requirements
Node.js >= 18
Claude Code CLI installed and authenticated
License
MIT
Available Tools
4 toolsclear_advisor_logClear Advisor LogADestructive
Clear the consultation log and metadata to start fresh for this project.
Inputs: no named parameters.
consult_opusConsult Opus AdvisorARead-only
Consult Claude Opus 4.7 for strategic advice. Opus runs via the Claude Code CLI with your existing subscription — no API key needed. The advisor maintains a per-project consultation log for continuity across calls. History is capped by both entry count (5) and token budget (~6K tokens) to prevent context bloat. Use this for architecture decisions, complex debugging, code review, or any problem that benefits from deeper reasoning.
Inputs: question (required), context, effort, files, +1 more.
read_advisor_logRead Advisor LogARead-only
Read the consultation log from prior Opus advisor calls for this project. Useful for reviewing past advice or getting context on decisions already made.
Inputs: last_n.
read_advisor_metaRead Advisor MetadataARead-only
Read structured metadata (latency, token counts, effort levels) from all consultations. Useful for understanding cost and performance patterns.
Inputs: last_n.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
Each tool has a uniquely defined purpose: clearing the log, consulting Opus, reading the log, and reading metadata. No two tools overlap in functionality, so an agent can easily distinguish them.
All tools follow a consistent verb_noun pattern using snake_case (e.g., consult_opus, read_advisor_log). The naming convention is uniform across the entire set.
Four tools is a well-scoped count for an advisor MCP server. Each tool addresses a distinct need (consultation, log management, metadata access) without unnecessary extras.
The tool set covers the core operations: consultation, log reading/clearing, and metadata inspection. A minor gap is the lack of a configuration tool to adjust Opus parameters, but the current surface is sufficient for most workflows.
Maintenance
Related MCP Connectors
Augments MCP Server - A comprehensive framework documentation provider for Claude Code
Shared memory MCP server for Claude Code, Codex, Gemini CLI and Cursor. Cited, editable.
Hosted X Ads MCP server for Claude, ChatGPT and Cursor, with each action switched per account.
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
Related MCP Servers
- AlicenseAqualityBmaintenanceAn MCP server that lets Claude Code consult stronger AI models (o3, Gemini 2.5 Pro, DeepSeek Reasoner) when you need deeper analysis on complex problems.1102 npm148MIT
- AlicenseAqualityDmaintenanceMCP server that lets Claude Code ask GPT Codex for adversarial planning, code review, debugging, research, and risk triage without leaving your project workflow.99 npm2MIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server that allows Claude Code to interact with the OpenAI Codex CLI.40 npm20MIT
- AlicenseNot gradedqualityBmaintenanceMCP server enabling Claude to consult Codex (GPT-5.x) mid-task for second opinions, plan/diff review, brainstorming, and codebase exploration via structured debates and permission-controlled interactions.2MIT