Fugu

Organization

aigorahub

fugu

Run a bounded Sakana Fugu repository task or explicit review through codex-fugu. Use when the user types /fugu, asks to use Fugu, or asks for a Fugu review.

Overview

Publisher	aigorahub
Repository	`elves`
Skill name	`fugu`
Stars	221
Forks	15
Bundled files	Instructions only
License	MIT
Links	Skill source Repository Homepage

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
Self-contained
Everything the model needs lives in the instructions — no extra files to sync.
Open source
Published by aigorahub on GitHub. Read the source before you install it.

Installation

Install the Fugu AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/aigorahub/elves/tree/main/aliases/claude/fugu
Enable the skill in any chat where you want it available.

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/aigorahub/elves.git /tmp/elves
mkdir -p .claude/skills
cp -r /tmp/elves/aliases/claude/fugu .claude/skills/fugu

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Fugu in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Fugu on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Fugu is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Fugu Planning or Review

Load the installed elves skill's Provider shortcut protocols and references/provider-shortcuts.md. Resolve scripts/run_fugu.sh from the active Elves skill root, keep the target repository as the working directory, pass through the validated mode/profile/context, and run it.

Host Fugu routing (required when the user omits a profile flag). Natural language "use Fugu" or plain /fugu <task> uses bare fugu/high by default. Before launch, state one short Fugu route: … line. Explicit user flags always win. Profile locks model + effort.

Host-native first if rg/git/gh can finish in under a minute.
Task mode: planning (default) vs review <scope> when the user asked for a review or audit.
Profile: use plain fugu/high by default. Use --deep only when regular Fugu needs xhigh effort. The host may select --cyber only for explicit security review or threat-model intent after a successful Cyber call in the current session. Only a user-explicit --cyber request may establish that proof. Otherwise, use regular Fugu. The user must explicitly select --ultra or --max.
Write: Fugu is read-only. The runner rejects --write.
Context: the isolation snapshot is always on. Put goal, paths, done-when, and out-of-scope in the task string. Add exact --include PATH only for non-gitignored files; if any include, run --preflight first and launch only when admitted. No separate "minimal snapshot" product.
Capture: redirect to a log file (never | tail / | head). Chat cancel does not stop the provider; wait up to the wall or kill the process group. On timeout/crash, harvest any Fugu partial salvage markers before relaunch. Verify findings host-native and clean up leftover process groups (see references/fugu-calling-guide.md).

Full decision table, route templates, wait/poll contract, and field notes: references/provider-shortcuts.md (Host routing when the user says "use Fugu") and references/fugu-calling-guide.md.

Plain /fugu <task> is a read-only planning task whose answer follows the request; /fugu review <scope> is the opinionated read-only review. Exact includes must be admitted and copied; both .env.* and *.env names are excluded. Live writable-state limits tolerate benign disappearing temporary subtrees and fail closed on other audit errors. macOS read-only cleanup remains best-effort and never claims recursive containment. The Linux lane omits procfs and exposes only a synthetic /proc/self/exe link to the qualified real Codex binary.

The runner uses the official codex-fugu launcher with policy-admitted tracked and non-ignored untracked context, closed interactive input, and a hard wall-clock bound. It selects regular fugu/high when the host chooses plain, fugu/xhigh with --deep, fugu-cyber/xhigh with --cyber, fugu-ultra-v2.0/high with --ultra (then fugu-ultra, then fugu-ultra-v1.1), or fugu-ultra-v1.1/max with --max. Regular/deep sessions are ephemeral; Ultra uses exact-session staged synthesis, with its state confined to the disposable isolated lane, events carried by a bounded host-owned pipe, final output pinned to a no-follow descriptor, and a final descriptor-safe writable-state audit after each settled phase. Do not replace it with an improvised API request or remove its sandbox and timeout controls.

Read-only review snapshots omit oversized binary media instead of failing the whole review. Video, audio, presentation, archive, image, font, and 3D binaries above the per-file limit are left out and listed in the context manifest with path, byte size, and reason; the 16 MiB per-file limit is not raised. Source, prose instructions, executable agent configuration, and --include paths still fail closed and ask for a derived text, image, or transcript artifact.

Fugu is optional. On any non-zero exit the runner prints one directive line naming the reason. A quota, authentication, catalog, runner, timeout, or provider failure all mean the same thing. Select another available independent reviewer instead of stopping, record requested route, actual route, and fallback reason, and do not claim a review ran when it did not: python3 "$ELVES_SKILL_ROOT/scripts/cobbler_agents.py" review-route --host claude-code --requested fugu --unavailable fugu=<reason> --json.

Frequently asked questions

What does the Fugu AI skill do?

Run a bounded Sakana Fugu repository task or explicit review through codex-fugu. Use when the user types /fugu, asks to use Fugu, or asks for a Fugu review.

Why use Fugu on TypingMind?

Because you install it once and use it with any model. Fugu is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Fugu in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/aigorahub/elves/tree/main/aliases/claude/fugu. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Fugu?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Fugu?

As many as you like. As long as a model supports skills, you can use Fugu with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Fugu AI skill free?

Yes. It is published on GitHub by aigorahub under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Change Impact

affaan-m

CommunityPopular

Perform evidence-backed impact analysis before changing code, data, interfaces, or governance documents, then compare the planned impact with the actual diff after implementation. Covers call paths,…

Test Collaboration

Inventory and govern a project's test assets. Connect requirements, business rules, risks, bugs, and cross-boundary contracts to stable TEST-IDs and executable evidence in the repository's existing…

Product Capability

Translate PRD intent, roadmap asks, or product discussions into an implementation-ready capability plan that exposes constraints, invariants, interfaces, and unresolved decisions before multi-service…

Django Patterns

Patrones de arquitectura Django, diseño de API REST con DRF, buenas prácticas de ORM, caché, señales, middleware y aplicaciones Django de nivel producción.

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Fugu Planning or Review

Frequently asked questions

Related AI Skills

Set up your own AI workspace now