magician

From idea to merged PR — autonomously, grounded in your code, gated only where it matters.

_{25 skills · a local code knowledge-graph · cross-session memory · parallel agent orchestration · an absolute destructive-command guard · zero required network calls or dependencies}

✦ What it is

Most AI coding tools make you describe the stack, pick templates, and babysit context. magician inspects your project on every session start, assembles targeted knowledge for each technology it finds, grounds itself in a local graph of your code, and runs the whole software lifecycle — design → plan → build → verify → review → ship — pausing only at the decisions that are genuinely yours.

/manifest

/transmute

⚡ The flow

Approve the plan once — then it executes autonomously, re-gating only on real side effects (writes to shared state, commits, push, PRs, deploys). Reads, searches, tests, and knowledge-graph lookups never interrupt you.

The manifest flow — full autonomous SDLC

flowchart TD
    A["/manifest"] --> B{"scope OK?"}
    B -- too large --> C["decompose into sub-projects"]
    B -- ok --> D["/conjure — design dialogue"]
    D --> E["approved spec"]
    E --> F["/blueprint — plan + parallelism map"]
    F --> G["/portal — git worktree isolation"]
    G --> H["/orchestrate — parallel agents"]
    H --> I["/ward — TDD throughout"]
    I --> J["/certify — tests + browser"]
    J --> K{all green?}
    K -- no --> H
    K -- yes --> L["/scrutinize — review + remediate"]
    L --> N["/seal — PR + loop until merged"]
    style A fill:#6c63ff,color:#fff
    style D fill:#6c63ff,color:#fff
    style F fill:#6c63ff,color:#fff
    style H fill:#43e97b,color:#000
    style I fill:#43e97b,color:#000
    style J fill:#43e97b,color:#000
    style L fill:#43e97b,color:#000
    style N fill:#4facfe,color:#000

Human gates (4 only): scope confirm → spec approval → plan approval → ship. Everything else: autonomous.

Dynamic project inspector — no manual stack selection

flowchart LR
    A["session start"] --> B["scan project files"]
    B --> C{"detect markers"}
    C --> D["package.json · tsconfig"]
    C --> E["pom.xml · *.xcodeproj"]
    C --> F["go.mod · Cargo.toml · pyproject"]
    C --> G["pubspec.yaml · project.godot"]
    D --> L["assign archetype + inject context"]
    E --> L
    F --> L
    G --> L
    L --> M["session ready in < 2s"]
    style A fill:#0d1117,color:#ccc,stroke:#555
    style M fill:#43e97b,color:#000

Polyglot stacks (Next.js + FastAPI + Go) get full coverage automatically — no pack selection.

Self-learning — intelligence grows each session

flowchart TD
    A["session ends"] --> B["Stop hook: chronicle"]
    B --> C["git log + diff (observable only)"]
    C --> D["write chronicle entry"]
    D --> E{"pattern seen 3x?"}
    E -- yes --> G["offer: create skill via /inscribe"]
    E -- no --> I["next session"]
    G --> I
    I --> K["load recent entries as context —\ncumulative intelligence without replay"]
    style B fill:#f7971e,color:#000
    style D fill:#f7971e,color:#000
    style G fill:#43e97b,color:#000
    style K fill:#6c63ff,color:#fff

🧠 What makes it different

🔒 Safety — an absolute destructive-command guard  _{new in 4.6.0}

Security is infrastructure, not advice.

A PreToolUse(Bash|PowerShell) hook inspects every shell command and unconditionally blocks the catastrophic ones. Because it exits before Claude Code evaluates permission rules, the block overrides allow rules and fires in every mode — default · acceptEdits · auto · bypass — with no escape hatch by design.

Wrappers (sudo, env, timeout, …) and sh -c '…' payloads are unwrapped first; a dangerous command merely named in a quoted argument (like a commit message) is not mistaken for execution. Honest scope: a denylist can't catch every obfuscation, so this is a deterministic floor layered under OS sandboxing, auto-mode's classifier, and model judgment — not a complete sandbox. Verified against a 90-case block/allow matrix.

🛠 Skills

25 skills, each with modern frontmatter (allowed-tools · disable-model-invocation · argument-hint · context: fork) that scales reasoning effort to the task. Approval gates use the structured AskUserQuestion tool, not prose.

🚀 Install

/plugin marketplace add https://github.com/Alexander-Tyagunov/magician
/plugin install magician@magician

codex plugin marketplace add Alexander-Tyagunov/magician

_{Codex loads adapter skills from .codex-plugin/skills/, preserving magician's gates while mapping to Codex tools. If it doesn't appear, add [plugins."magician@magician"]\nenabled = true to ~/.codex/config.toml.}

🗂 Workspace — team memory

.workspace/
├── shared/         ← git-committed (whole team)
│   ├── specs/       design specs   (/conjure)
│   ├── plans/       impl plans      (/blueprint)
│   ├── research/    findings        (/magic)
│   ├── decisions/   ADRs
│   └── postmortems/ (/autopsy)
└── local/          ← always gitignored
    ├── prefs.md     personal prefs
    └── session.md   pre-compaction state

🧰 Bundled CLIs  _{(on PATH when the plugin is enabled)}

❤ Support this work

If magician saves you time, consider sponsoring — it funds new skills, broader framework lore, and community support.

❤ Sponsor on GitHub →

_{MIT © Alexander Tyagunov · built for Claude Code & Codex}