Install and connect WhiteMagic.

Find the WhiteMagic MCP server (whitemagic.dev; official MCP registry id io.github.lbailey94/whitemagic-mcp). Ask me which release channel I want; do not install the experimental preview unless I explicitly choose it. Stable Gen 2 is the default.
Choose the install command matching my operating system:
Linux (POSIX shell): curl -fsSL "https://www.whitemagic.dev/install.sh?ref=ask-ai" | sh -s -- --version v9.3.4 --dir "$HOME/.local/bin" --ref ask-ai
macOS (POSIX shell): curl -fsSL "https://www.whitemagic.dev/install.sh?ref=ask-ai" | sh -s -- --version v9.3.4 --dir "$HOME/.local/bin" --ref ask-ai
Windows (PowerShell): & ([scriptblock]::Create((irm 'https://raw.githubusercontent.com/lbailey94/whitemagic/v9.3.4/scripts/install.ps1'))) -Version 'v9.3.4' -Dir (Join-Path $env:LOCALAPPDATA 'WhiteMagic\bin') -Ref 'ask-ai'
Do not run a POSIX curl command in PowerShell. If a platform/channel has no published target, stop and explain that rather than substituting another binary.
Only if I explicitly choose experimental Gen 3, use the current supported Linux or Apple Silicon macOS command:
Linux: curl -fsSL "https://www.whitemagic.dev/install.sh?ref=ask-ai" | WM_STORE="$HOME/.local/share/whitemagic/gen3-preview" sh -s -- --version v10.3 --dir "$HOME/.local/bin/whitemagic-gen3" --ref ask-ai --no-path
macOS: curl -fsSL "https://www.whitemagic.dev/install.sh?ref=ask-ai" | WM_STORE="$HOME/.local/share/whitemagic/gen3-preview" sh -s -- --version v10.3 --dir "$HOME/.local/bin/whitemagic-gen3" --ref ask-ai --no-path
Gen 3 installs into $HOME/.local/bin/whitemagic-gen3 with --no-path and uses a separate store at $HOME/.local/share/whitemagic/gen3-preview. It has no Windows or Intel Mac target in the current channel record.
The numbered readiness and client-write workflow below is for Gen 2 only. Do not run Gen 2's `wm grimoire`, `wm connect`, `wm setup`, backup, or update workflow against Gen 3. For a preview MCP server, inspect that installed binary's `wm --help` and `wm serve --help`; use its full absolute binary path and pass the separate store before `serve`. MCP clients do not expand shell variables in JSON: replace the illustrative home prefix with your expanded absolute path. Example shape: command "/absolute/home/.local/bin/whitemagic-gen3/wm", args ["--store", "/absolute/home/.local/share/whitemagic/gen3-preview", "serve"]. Show me the proposed client configuration before writing it; apply configuration changes only after I approve them, and restart the client afterward.
For stable Gen 2, run `wm --version`, then follow the stable readiness steps below. Report which generation and version you chose and any diagnostic result; do not share private memory contents.

Who can do this

• ·Works in most clients that can run shell commands – Claude Code, Codex CLI, opencode, and similar terminal-capable agents can carry the prompt out end to end.
• ·Clients that can browse the official MCP registry can search for io.github.lbailey94/whitemagic-mcp and add it directly, with no shell step.
• ·Terminal-native path, one client at a time: wm setup <client> (claude · cursor · windsurf · codex · opencode) shows the change first; add --write to apply it with a timestamped backup.
• ·No terminal at all? Paste the matching config by hand from the client tabs below – the generic mcpServers entry works in most clients.
• ·Prefer not to install a binary? The npm launcher npx -y whitemagic-mcp@9.3.4 serve runs the stable Gen 2 server.

The per-client tabs below remain the manual fallback; where a client ships its own add command, that one-liner also works (Claude Code: claude mcp add whitemagic -- wm serve).

claude mcp add whitemagic -- wm serve

{
  "mcpServers": {
    "whitemagic": {
      "command": "wm",
      "args": ["serve"]
    }
  }
}

[mcp_servers.whitemagic]
command = "wm"
args = ["serve"]

{
  "mcpServers": {
    "whitemagic": {
      "command": "wm",
      "args": ["serve"]
    }
  }
}

{
  "mcpServers": {
    "whitemagic": {
      "command": "wm",
      "args": ["serve"]
    }
  }
}

{
  "servers": {
    "whitemagic": {
      "type": "stdio",
      "command": "wm",
      "args": ["serve"]
    }
  }
}

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "whitemagic": {
      "type": "local",
      "command": ["wm", "serve"],
      "enabled": true
    }
  }
}

{
  "mcpServers": {
    "whitemagic": {
      "command": "wm",
      "args": ["serve"]
    }
  }
}

{
  "mcpServers": {
    "whitemagic": {
      "command": "wm",
      "args": ["serve"]
    }
  }
}

Run once, then ask your client what it remembers.

wm ingest reads files from your disk, chunks them, and writes them into a store with provenance tags. It is idempotent: a per-file SHA-256 ledger at <store>/ingest_ledger.jsonl means re-runs are no-ops for unchanged files, and an interrupted ingest resumes by running the same command again.

wm ingest --source ~/Notes --galaxy research --redact --dry-run

wm ingest --source ~/Notes --galaxy research --redact

ingest flags

--source <dir>

Folder to walk (required). The walk is recursive and skips build/VCS directories (.git, node_modules, target, .next, …).

--galaxy <name>

Target galaxy. Default: research for documents, sessions for transcripts.

--redact

Scrub credential-shaped spans (PEM keys, prefixed tokens, assignment values) and ingest the file instead of skipping it.

--dry-run

Report without writing or creating the store.

--limit <n>

Only the first N files in walk order.

--include-credential-files

Ingest files whose name looks credential-bearing (secrets.txt); requires --redact, so content is scrubbed before storage.

--store <path> · --wait <n>

Point at a specific store; wait up to N seconds for a busy (live-server) store before failing.

Scan ~/Notes recursively into a research galaxy with `wm ingest --source ~/Notes --galaxy research --redact`, redacting credential-shaped spans. Dry-run first, then run it for real, and report the files ingested and skipped. Everything stays on this machine.

Beyond folders

• ·Direct memories – memory.create for one memory, memory.batch_create for a batch, session.record for conversation turns, all via the wm meta-tool. Tag cross-store imports with project:<name> so their origin stays attributable.
• ·Move galaxies – galaxy.export writes a galaxy as JSON; galaxy.import loads it into another store or machine.
• ·Savings ledger – `wm ledger` reports what stays out of model context: bytes stored vs injected by continuity and recall, plus the local operation mix. Local-only; `--json` for machines, `--calibrate` sets a per-store bytes/token divisor.
• ·Per-project stores – one store per project. Point the project's client config at its own directory and set WM_PROJECT=<name> so the handshake names the scope.