How do I use fugu-mcp?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@fugu-mcp Ask Fugu to review this code for potential bugs." 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.

fugu-mcp

by jamesaoyama

Overview Schema Source Related Servers Score Discussions Admin

TypeScript

Remote

fugu-mcp

An MCP server that exposes Sakana Fugu — an OpenAI-compatible orchestration LLM — as on-demand tools inside Claude Code. Use it to pull a second opinion or run a small multi-model panel on a discrete, self-contained subproblem, without leaving your Claude Code session.

Fugu is a tool, not the driver. This does not make Fugu the main model for Claude Code. Claude Code only drives on Anthropic / Bedrock / Vertex / Foundry models. fugu-mcp lets the Claude main loop call out to Fugu on demand — it is a narrow tool, not a model swap.

What you get

Two tools, both backed by Sakana's POST /v1/chat/completions endpoint:

Tool	Use it for
`ask_fugu`	A general second opinion on a hard, self-contained question (tricky algorithm, design trade-off, "is my approach sound?").
`fugu_second_opinion`	A review-shaped critique: hand it `content` (code, diff, answer, plan) plus a `question`, get back a skeptical senior-reviewer pass.

Plus a /fugu slash command for manual invocation.

Related MCP server: codex-cli-mcp-tool

Install

Requires Node 20+ (Node 26 recommended; uses native fetch).

git clone https://github.com/<your-org>/fugu-mcp.git
cd fugu-mcp
npm install
npm run build            # tsc -> dist/

export SAKANA_API_KEY=...   # your real key; never commit it

# run from the repo root so $(pwd) resolves to your clone
claude mcp add fugu -- node "$(pwd)/dist/index.js"

Then in Claude Code, e.g.: "Ask Fugu to cross-check this approach."

claude mcp add inherits your shell environment, so the exported SAKANA_API_KEY (and any FUGU_* overrides) flow through to the server.

Alternative: `.mcp.json`

To check the registration into a project (so collaborators get it automatically), add a .mcp.json at the repo root instead of running claude mcp add. A ready-to-edit template lives at .mcp.json.example:

{
  "mcpServers": {
    "fugu": {
      "command": "node",
      "args": ["dist/index.js"],
      "env": {
        "SAKANA_API_KEY": "REPLACE_WITH_YOUR_SAKANA_API_KEY",
        "FUGU_BASE_URL": "https://api.sakana.ai/v1",
        "FUGU_DEFAULT_MODEL": "fugu-ultra",
        "FUGU_TIMEOUT_MS": "120000"
      }
    }
  }
}

Do not commit a real SAKANA_API_KEY in .mcp.json. Either keep .mcp.json out of version control, or leave the placeholder and set the real key via the environment (the server reads process.env.SAKANA_API_KEY regardless of where the session is launched).
args uses a relative dist/index.js; this resolves against the directory Claude Code is launched from. Use an absolute path (/path/to/fugu-mcp/dist/index.js) if you run sessions from elsewhere.

Use with OpenClaw

OpenClaw supports MCP natively, so the same server works there — no separate build. After npm run build:

openclaw mcp add fugu \
  --command node \
  --arg "/path/to/fugu-mcp/dist/index.js" \
  --env SAKANA_API_KEY=your_key_here

openclaw mcp add probes the server before saving. Confirm with openclaw mcp list, and the ask_fugu / fugu_second_opinion tools become available to your OpenClaw agents. Run openclaw mcp reload after changing config so the new settings take effect on the next turn.

Tools

`ask_fugu`

A general second opinion / hard self-contained query. The prompt must contain everything Fugu needs — it cannot see your repo, files, or the current conversation.

Argument	Type	Required	Default	Notes
`prompt`	string	yes	—	The full, self-contained problem or question.
`system_prompt`	string	no	—	Steer Fugu's role/behavior.
`model`	`"fugu"` \| `"fugu-ultra"`	no	`FUGU_DEFAULT_MODEL`	Pick a specific model.
`max_tokens`	integer > 0	no	`2000`	Max output tokens.

Example invocation:

{
  "name": "ask_fugu",
  "arguments": {
    "prompt": "I'm choosing between a single-writer WAL and a per-shard log for a 50k-writes/sec ledger. Walk through the durability/throughput trade-offs and recommend one. Assume NVMe and fsync-per-commit.",
    "system_prompt": "You are a senior distributed-systems engineer.",
    "model": "fugu-ultra",
    "max_tokens": 1500
  }
}

In conversation you can just say: "Use ask_fugu to sanity-check this proof."

`fugu_second_opinion`

A review-shaped critique. Wraps a skeptical-senior-reviewer system prompt internally, so you only supply the material and the question.

Argument	Type	Required	Default	Notes
`content`	string	yes	—	The material to review: code, a diff, an answer, a design, or a plan.
`question`	string	yes	—	What you want assessed.
`model`	`"fugu"` \| `"fugu-ultra"`	no	`FUGU_DEFAULT_MODEL`	Pick a specific model.

Example invocation:

{
  "name": "fugu_second_opinion",
  "arguments": {
    "content": "func transfer(from, to *Account, amt int) {\n  from.mu.Lock(); to.mu.Lock()\n  from.bal -= amt; to.bal += amt\n  from.mu.Unlock(); to.mu.Unlock()\n}",
    "question": "Is this concurrency-safe?",
    "model": "fugu-ultra"
  }
}

In conversation: "Get a fugu_second_opinion on this diff — is the error handling complete?"

`/fugu` slash command

/fugu <your question> routes your text to ask_fugu and asks Claude to summarize Fugu's answer and note where it agrees or disagrees — treating Fugu as one voice in a panel, not ground truth. See .claude/commands/fugu.md.

Configuration (environment variables)

All config is via env. Copy .env.example to .env for local use (.env is gitignored).

Variable	Required	Default	Purpose
`SAKANA_API_KEY`	yes (at call time)	—	Sent as `Authorization: Bearer <key>`. Optional at startup so the server can boot and answer `tools/list`; a missing key surfaces as a clear tool error only when a tool is actually called — it never crashes the process. Env only; never write a real key to a tracked file.
`FUGU_BASE_URL`	no	`https://api.sakana.ai/v1`	OpenAI-compatible base URL. Requests go to `<base>/chat/completions`.
`FUGU_DEFAULT_MODEL`	no	`fugu-ultra`	Model used when a tool call omits `model`.
`FUGU_TIMEOUT_MS`	no	`120000`	Per-request timeout in milliseconds. Orchestrators are slow; the default is generous.

A malformed FUGU_BASE_URL (not a URL) or non-positive FUGU_TIMEOUT_MS fails fast at startup with a clear message. A missing key does not.

Hooks (for working on this repo)

This repo ships .claude/settings.json with hooks that keep the build green and guard secrets while you edit fugu-mcp itself:

PreToolUse on Write | Edit | Bash → runs scripts/guard-secrets.sh, which scans the payload for an API-key pattern and blocks the operation if a real-looking key is detected (so a key can't be written into a file or slipped into a commit).
PostToolUse on Edit | Write | MultiEdit → runs npm run typecheck (non-blocking; keeps types green during edits).

These hooks are for development of fugu-mcp; they are not required to use the MCP server in another project.

`MCP_TIMEOUT` for the consuming session

Fugu calls can take tens of seconds (up to ~120s). Claude Code's default MCP startup/tool timeout can be too short. In the session that consumes this server, set:

export MCP_TIMEOUT=120000

This raises Claude Code's MCP timeout to match the server's own FUGU_TIMEOUT_MS default, so slow Fugu responses aren't cut off.

Standalone vs integrated

Use Fugu standalone when: you want a one-shot answer and you're happy outside Claude Code. Drive Sakana's OpenAI-compatible API directly from Codex, Cursor, or curl. That path is simpler and avoids a hop.

Use this MCP tool when: you're already mid-task in Claude Code and want a narrow second opinion — a different model's take on a discrete, self-contained subproblem — without breaking flow. fugu-mcp is deliberately small: two tools, second-opinion shaped, no repo access.

Trade-offs to keep in mind:

Latency. Going through the Claude orchestrator adds a hop, and Fugu itself is slow (tens of seconds, occasionally up to ~120s). It is not for tight iteration loops.
Cost. Fugu output is roughly ~$30 / 1M output tokens. max_tokens defaults are kept modest (ask_fugu defaults to 2000); keep prompts and outputs tight.
Region. Sakana's API is blocked in the EU/EEA. If you're in that region you won't be able to call it.
Model IDs. This server defaults to fugu-ultra and exposes fugu / fugu-ultra. Verify the exact, current Fugu model IDs and endpoint against the Sakana docs — they can change. If they differ, it's a config edit (FUGU_DEFAULT_MODEL, FUGU_BASE_URL), not a code change.

Development

npm run build        # compile TypeScript to dist/
npm run typecheck    # tsc --noEmit
npm test             # vitest run (mocked fetch; no network, no key needed)
npm start            # node dist/index.js (stdio server)

Available Tools

2 tools

ask_fuguAsk Fugu (second opinion)A

Ask Sakana Fugu (a separate orchestration LLM) for a second opinion on a HARD, SELF-CONTAINED problem, or to cross-check your own reasoning with a different model. Good for: a tricky algorithm/proof, a thorny design trade-off, a 'is my approach sound?' gut-check, or a multi-model panel on a discrete question. The prompt MUST contain all needed context — Fugu cannot see the repo, files, or this conversation. Do NOT use it for interactive or iterative repo work, multi-step edits, running commands, or anything that needs to read/modify local files — do that yourself. Calls can be slow (tens of seconds).

ParametersJSON Schema

Name	Required	Description
`prompt`	Yes	The full, self-contained problem or question to send to Fugu. Include all context Fugu needs — it cannot see your repo or conversation.
`system_prompt`	No	Optional system prompt to steer Fugu's role/behavior (e.g. 'You are a senior distributed-systems engineer.').
`model`	No	Fugu model id. Omit to use the server's default (FUGU_DEFAULT_MODEL).
`max_tokens`	No	Max output tokens. Defaults to 2000.

TDQS

A4.5/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations given, so description handles all burden. Discloses that Fugu is a separate LLM, cannot see repo/files/conversation, and calls can be slow (tens of seconds). Does not mention read-only/destructive but it's implied as a LLM call. Adequate for safe use.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two well-structured paragraphs, front-loaded with purpose and key constraints. Every sentence adds value; no fluff. Efficiently conveys necessary information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Despite no output schema and no annotations, the description covers purpose, usage guidelines, parameter requirements, behavioral traits (slowness, self-contained), and limitations. Complete for an agent to select and invoke correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% with descriptive parameter definitions. Description adds minor context (e.g., 'Include all context Fugu needs') but largely overlaps with schema. At baseline of 3 for high coverage; no significant value beyond schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Title and description clearly state it's for a second opinion from a separate LLM. Specifies exact use cases (hard self-contained problems, tricky algorithms, design trade-offs) and distinguishes from sibling by explicitly stating what it is not for (interactive repo work, commands, etc.).

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly lists when to use (hard problems, gut-check) and when not (iterative repo work, multi-step edits). Provides context that prompt must be self-contained, and advises doing other tasks yourself. No ambiguity.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

fugu_second_opinionFugu second opinion (review)A

Get a rigorous SECOND OPINION from Sakana Fugu (a separate LLM) on a discrete piece of work. Provide the content to review (code, diff, answer, design, or plan) and a specific question. Fugu acts as a skeptical senior reviewer and returns concrete issues, risks, and improvements. Best for cross-checking your own output with a different model on a self-contained artifact. All needed context must be in content — Fugu cannot see the repo or this conversation. Do NOT use it for interactive/iterative repo work or anything needing local file access. Calls can be slow.

ParametersJSON Schema

Name	Required	Description
`content`	Yes	The material to review: code, a diff, a written answer, a design, or a plan. Include everything Fugu needs to judge it — it has no other context.
`question`	Yes	What you want Fugu to assess (e.g. 'Is this concurrency-safe?', 'Does this proof hold?', 'Is this API design sound?').
`model`	No	Fugu model id. Omit to use the server's default (FUGU_DEFAULT_MODEL).

TDQS

A4.4/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden. It discloses that Fugu is a separate LLM, acts as a skeptical senior reviewer, and returns 'concrete issues, risks, and improvements.' It also notes that Fugu cannot see the repo or conversation, and that all context must be in 'content.' While it could mention potential timeouts or output format, the description provides sufficient behavioral context.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise (3-4 sentences) and front-loaded with the core purpose. Every sentence adds value: purpose, best use case, constraint, warning, and performance note. No unnecessary words.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the absence of an output schema, the description explains return values as 'concrete issues, risks, and improvements.' This is sufficient but could be more specific (e.g., format). The tool has three well-documented parameters and clear usage guidance, leaving minimal gaps. A minor omission is lack of mention of rate limits or output length.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with each parameter having a detailed description in the schema (e.g., 'content' includes examples of what to review). The tool description does not add additional parameter-level semantics beyond the schema, so the baseline score of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: 'Get a rigorous SECOND OPINION from Sakana Fugu on a discrete piece of work.' It specifies the verb (get), resource (second opinion), and scope (discrete piece), and differentiates from the sibling 'ask_fugu' by emphasizing cross-checking and skepticism.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit guidance: 'Best for cross-checking your own output with a different model on a self-contained artifact.' It also states when not to use it: 'Do NOT use it for interactive/iterative repo work or anything needing local file access.' Additionally, it mentions performance ('Calls can be slow'), giving clear context for use.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

2 tool updatesv0.1.0Jun 23, 2026
- First observedask_fugu
- First observedfugu_second_opinion

Install Server

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A3.7/5.0

Scored 2026-06-23 21:30 across 2 tools

Disambiguation2/5

Both tools serve essentially the same purpose—getting a second opinion from Sakana Fugu. The descriptions differ slightly in phrasing but offer the same core functionality, making it unclear when to use one versus the other.

Naming Consistency2/5

Tool names are inconsistent: 'ask_fugu' follows a verb_noun pattern, while 'fugu_second_opinion' uses a noun_adjective pattern. Mixed naming conventions reduce predictability.

Tool Count2/5

With only two tools covering a single niche (second opinions), the set feels unnecessarily thin. The functionality could easily be served by one well-designed tool.

Completeness1/5

The tool surface is severely incomplete for any meaningful server scope. Not only is there no variety of operations, but the two tools overlap almost entirely, leaving no discernible domain coverage beyond a single redundant capability.

Maintenance

ActivityStale

ResponsivenessNo issues

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

mcp
dev.augments
Augments MCP Server - A comprehensive framework documentation provider for Claude Code
Developer Tools Documentation Access
2128
Remote Desktop CommanderOAuth
app.desktopcommander
Hosted MCP server connecting claude.ai, ChatGPT and other AI apps to your own computer
Shell Access OS Automation
219
gread A
io.github.NitroRCr
An MCP server that gives your AI access to the source code and docs of all public github repos
Version Control Code Analysis Documentation Access
5361
strata
ai.klavis
MCP server for progressive tool usage at any scale (see https://klavis.ai)
Agent Orchestration Autonomous Agents Developer Tools
25,748

Related MCP Servers

mcp-agent-review
Code Analysis AI & Machine Learning
lzx1413
A
license
Not graded
quality
C
maintenance
An MCP server that provides agentic code review powered by OpenAI-compatible models, designed for use with Claude Code.
Updated 4 months ago (2026-04-29 01:52 UTC)
1
MIT
codex-cli-mcp-tool
Code Execution Coding Agents
Mr-Tomahawk
A
license
Not graded
quality
F
maintenance
An MCP server that allows Claude Code to interact with the OpenAI Codex CLI.
Updated a year ago (2025-08-19 07:26 UTC)
21
21
MIT
codex-fusion-mcp
AI & Machine Learning Developer Tools Code Analysis
Tritlo
A
license
Not graded
quality
B
maintenance
MCP 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.
Updated 2 months ago (2026-07-06 16:11 UTC)
2
MIT
ollama-code-mcp
Coding Agents Developer Tools AI & Machine Learning
darthzen
A
license
A
quality
B
maintenance
An MCP server that lets Claude Code delegate coding tasks such as code generation, review, refactoring, and test writing to a local Ollama instance running a Qwen3 model.
Updated a month ago (2026-07-29 21:52 UTC)
9
MIT

fugu-mcp

What you get

Install

Alternative: .mcp.json

Use with OpenClaw

Tools

ask_fugu

fugu_second_opinion

/fugu slash command

Configuration (environment variables)

Hooks (for working on this repo)

MCP_TIMEOUT for the consuming session

Standalone vs integrated

Development

Available Tools

TDQS

TDQS

Tool Schema Changelog

TDQS

Maintenance

Resources

Looking for Admin?

Related MCP Connectors

Related MCP Servers

mcp-agent-review

codex-cli-mcp-tool

codex-fusion-mcp

ollama-code-mcp

Alternative: `.mcp.json`

`ask_fugu`

`fugu_second_opinion`

`/fugu` slash command

`MCP_TIMEOUT` for the consuming session