A small, vetted, self-evolving harness for Claude Code — curated, not dumped.
🇷🇺 Кратко на русском #
Небольшой курированный харнесс для Claude Code — не свалка из сотен блоков, а горстка скиллов, которые заслужили место, плюс каркас безопасности и (главное) цикл самоулучшения: правки пользователя становятся памятью, находки — скиллами, вход нового инструмента идёт через гейт /vet
(безопасность — блокирующий шаг). Никаких авто-хуков, которые мутируют что-то без спроса. Подробности — в docs/EVOLUTION.md. Полный текст ниже на английском.
Most "everything for Claude Code" collections optimize for count — hundreds of agents, skills, rules, and event hooks that auto-run on your tool calls. It looks impressive and fails quietly: bloated context, overlapping triggers fighting to activate, and hooks that mutate things without asking.
This repo optimizes for the opposite. A handful of skills that earn their place, a spine of safety guards, and — the actual product — a system that improves itself over time, safely.
A harness that grows is worth more than a harness that's big.The thing that compounds isn't the block count — it's the intake gate, the hygiene loop, and the learning machinery. →is the heart of this repo; start there.docs/EVOLUTION.md
- ✅ A curated set of executable skills + always-on principles you can drop into
~/.claude/
. - ✅ A
method(/vet
) for deciding what to adopt, and a registry so you never evaluate the same tool twice. - ✅ A self-evolution loop— corrections become memory, findings become skills, the stack stays clean. - ✅ Safety by design— no surprise autonomous actions. - ❌ Not a 500-block mega-bundle. Not a framework you fork. No auto-running hooks.
Single-skill self-learning harvesters are trending, and the good ones are genuinely good. This repo is the system that pattern lives in: four learning layers (corrections → memory → skills → gated intake) instead of one harvest step, a promotion gate so unverified guesses never become authoritative skills, an intake gate (/vet
) so the stack doesn't bloat as it learns, and a hygiene loop (system-health
) so it doesn't rot. If you only want the harvest step, Kulaxyz/self-learning-skills does it well — we credit and cherry-pick it in
docs/EVOLUTION.md
operating discipline, start here.
Vet before adopt. Nothing enters on hype. Every incoming tool runs through one pipeline — relevance →security as a blocking gate→ duplication → real ROI → verdict. The security phase reads a skill's body as abehavioralinstruction and checks the agentic risks (hidden instructions, prompt injection, trigger abuse, excessive agency) a code-only scan misses.It evolves. A user correction is the strongest learning signal; a finding is captured at discovery; a 5+-iteration procedure becomes a skill — but only past apromotion gate(verified check + named failure pattern + a ruled-out dead-end), so confident guesses never get enshrined as truth. The system compounds instead of going stale. →docs/EVOLUTION.md
Cherry-pick over giant. A good idea inside a huge framework doesn't justify installing the framework. Extract the technique, leave the bloat.No auto-hooks. Auto-mutation without confirmation is exactly the failure mode to avoid. →rules/safety-guards.md
Full rationale: docs/METHODOLOGY.md.
| Skill | What it does |
|---|---|
vet |
workflow-upgrade
goal
konsilium
ship-secure
pre-push
system-health
teach
longread
Two by design, not sixty — each reflects the verify-before-trust stance, not filler.
| Agent | What it does |
|---|---|
adversarial-verifier |
tool-scout
/vet
— reads a tool's repo/docs and returns a structured security report; never installs or runs.| Rule | What it covers |
|---|---|
code-principles.md |
safety-guards.md
root-cause-discipline.md
new-project.md
detective-mindset.md
—docs/EVOLUTION.md
how the harness improves itself(start here).— why curated, not dumped.docs/METHODOLOGY.md
— a file-based, self-learning memory pattern.docs/MEMORY_SYSTEM.md
— task → tool routing, proactively.docs/TOOL_ROUTING.md
— backup → surgical → verify → record.docs/CHANGE_DISCIPLINE.md
— a registry template and a workedexamples/
/vet
verdict.
| I want to… | Use |
|---|---|
| Decide whether to adopt a new tool/skill/MCP | vet |
| Run a deliberate pass to improve my whole setup | workflow-upgrade |
| Hand a goal to the agent and get back a finished, verified result | goal |
| Pressure-test a high-stakes decision (go/no-go) | konsilium |
| Make sure a web app is safe before launch/handoff | ship-secure |
| Make sure nothing junk leaves on push/deploy | pre-push |
| Keep my Claude Code config from rotting | system-health |
| Actually understand a session/article/concept | teach |
| Write an analytical longform piece | longread |
Install manually — copy only what you want. (No curl | bash
: this harness practices what it preaches.)
git clone https://github.com/Sanexxxx777/curated-claude-code
cd curated-claude-code
cp -r skills/vet ~/.claude/skills/
cp -r skills/ship-secure ~/.claude/skills/
cp agents/*.md ~/.claude/agents/
Restart Claude Code, then invoke a skill by its name (e.g. /vet
).
This harness was built the way it tells you to build — by cherry-picking techniques and crediting them:
konsilium
adapts Andrej Karpathy's(independent opinions → anonymized peer review → chairman synthesis).llm-council
ship-secure
adapts Prajwal Tomar's "Vibe Coders Are Getting Sued" launch playbook.code-principles
is derived from Karpathy's widely-shared CLAUDE.md principles.goal
's design phase cherry-picks loop-design ideas from.ksimback/looper
teach
adapts Anthropic's guided-learning prompt.- The skill promotion gate in
EVOLUTION.md
cherry-picks the verification bar from.Kulaxyz/self-learning-skills
I build and harden Claude Code workflows, agentic setups, and web apps — and I do managed hosting and site builds. If these patterns are useful to your team, or you want help applying them, reach out:
GitHub: open an issue, or@Sanexxxx777Email:sanexxx777@gmail.com
PRs welcome — see CONTRIBUTING.md. The bar: it must pass its own
/vet
, ship no auto-hooks, contain no secrets, and stay surgical. Curation is the point; volume isn't.If this saved you time, a ⭐ helps others find it.
MIT © Aleksandr Shulgin (@Aleksandr_NFA) · @Sanexxxx777