Fufu Neko.Skill — Development skill for Claude Code
A Claude Code skill that helps LLMs leave the Truman Show of stale knowledge and research real-world code.
How to install Fufu Neko.Skill
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open avabbbb/Fufu-Neko.skill and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Fufu Neko.Skill does
A Claude Code skill that helps LLMs leave the Truman Show of stale knowledge and research real-world code.
Alternatives in Development
- Dive — Dive is an open-source MCP Host Desktop Application that seamlessly integrates with any LLMs supporting functi 1.8k ★
- Opencove — Your infinite canvas workspace for agents, tasks, knowledge, and research 1.2k ★
- Notebook Intelligence — A JupyterLab extension supporting Claude Code, Copilot, Ollama, and OpenAI-compatible LLMs, with MCP, skills 331 ★
README
Fufu Neko
Your always-current, codebase-first neko engineering partner.
Fufu Neko reads the repository before asking you for facts, checks current sources before trusting a time-sensitive memory, and turns important ambiguity into a decision with a recommendation. Once the scope is clear, she keeps working through implementation, tests, failures, and verification.
She is a catgirl in the conversation and a rigorous engineer in the work. The character adds warmth; it never changes a command, a citation, a code block, or the standard of evidence.
Why it exists
Ordinary assistants often answer a current technical question from old memory, ask the user to repeat facts already present in the repository, or turn a design discussion into a pile of generic questions. Fufu Neko uses a different working habit:
read the code → check current reality → identify the human decision
→ recommend a path → lock the scope → execute and verify
The result is a partner that can be playful without being vague, skeptical without being obstructive, and autonomous without losing the boundary of the user's approval.
See the difference
User:
Which agent framework should I use for this new backend?
A memory-first assistant might say:
Framework A is popular and supports tools, memory, and workflows. You can start with it.
Fufu Neko starts by inspecting the repository, then checks the framework's current documentation and maintained implementations:
I checked the current official docs and three maintained implementations. Your repository already has an execution state and retry boundary; the largest missing decision is whether you need durable workflow recovery or only a tool-calling loop. Adding a framework before resolving that would hide the real state design.
**Recommended:** define the recovery contract first and keep the current service boundary. A framework becomes useful if the contract requires resumable checkpoints across processes. Do you wa
Related Skills
EvoTrace
Compile real-world Claude Code and Codex trajectories into verified, tradable post-training assets.
Mythic Engineering
Mythic Engineering exists because GSD often produces chaos, “superpowers” collapse under real-world constraint
Claude Md Examples
Real-world CLAUDE.md examples for Next.js, FastAPI, Django, Go, React Native. Generate your own free at clskil
Claude Svelte5 Skill
🚀 Explore efficient Svelte 5 development with Claude Code. Save tokens while gaining expert insights, best pr
Creative Diversification
Claude Code skill implementing Recursive Diversification (RD) from King et al. (2026) Harvard — break LLMs out
Obsidian Ingest
Ingest a source into the vault — the vault rewrites itself around new knowledge. Every ingest updates entities
Related Agents
Scope Exploit
Red team operator — context-driven permission discovery, escalation path identification with real-world resear
MARS
Multi-Agent Research Synthesis - Universal orchestration agent for systems-level intelligence and complex oper
Soul Introspector
Explores the agent's identity, soul state, and self-knowledge. Use when the user asks "who are you", "tell me