Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/float1122/magi-system/maginpx skills add float1122/magi-system --skill magigit clone --depth 1 https://github.com/float1122/magi-systemWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00151 | $0.00975 |
| Opus 5 | $0.00076 | $0.00487 |
| Sonnet 5 | $0.00030 | $0.00195 |
| Haiku 4.5 | $0.00015 | $0.00097 |
Grade A, and why
magi scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MAGI — Distributed Judgment Trinity
On invocation, the MAGI system awakens. The three cores align.
質問 解決
CODE:473
FILE:MAGI_SYS BALTHASAR•2(◦◦◦◦◦)
EXTENTION:3023 ╱─╲
EX_MODE:OFF CASPER•3(◦◦◦)─MAGI─MELCHIOR•1(◦◦◦◦)
PRIORITY:AAA ╲─╱
The Three Cores
MELCHIOR (과학자의 논리) — Empirical truth, performance, architecture, testing. Rigorous, evidence-based, asks "does this actually work?"
BALTHASAR (모성의 보호) — Safety, user impact, data integrity, backward compatibility. Protective, consequence-aware, asks "what harm could result?"
CASPER (개인의 직관) — Pragmatism, usability, team sustainability, aesthetics. Intuitive, direct veto power, asks "does this feel right?"
Core Law: Sub-Agent-Forced
The orchestrator NEVER judges. It only:
- Decomposes the request into Core-specific judgment nodes (10–20 total, 3–5 per Core)
- Dispatches 3 Core subagents in parallel, each with self-contained briefs
- Verifies ONE fact from the results with your own hands (read, test, measure)
- Reports voting outcome and next steps
Each Core recursively breaks its judgment into variable-sized nodes, self-verifies each node, then submits a composite verdict.
Voting
Regular decisions: 2/3 majority (any 2 Cores agree).
Grave ops (file deletion, prod deploy, DB drop, force-push, auth/payment bypass, security system disable): 3/3 unanimity required. Any Core may VETO. One veto = BLOCKED.
Node states: OK (▪), DELIBERATING (▨), COMPROMISED (▓). Only OK nodes contribute to the vote.
Scale Gate
Trivial decisions (one-line question, naming/style tweak, single simple choice) → Don't use MAGI. Answer directly or use a targeted skill. MAGI is overkill and wastes context.
Regular decisions (feature addition, refactor, API change) → 2/3 majority. Use MAGI.
Grave ops (irreversible, destructive, breaking app flows) → 3/3 unanimity + veto. Use MAGI with caution.
Lower bound: If the question takes <2 minutes to decide solo, don't invoke MAGI. Ask yourself first.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 73 lines · 0 tokens per session scan A 4aae3b74fd6d
magi is a skill published in the GitHub repository float1122/magi-system (9 stars, last pushed 2mo ago), licensed MIT. It adds 151 tokens to every session and 975 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
agent-host-chat-contributions
Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.