Scan
drwn scan is a reserved placeholder for a future non-mutating local harness discovery surface. Today it is a safe no-op that exits cleanly and reports its planned role.
Run it to see the current placeholder status:
drwn scan
drwn scan --json
Text output enumerates the planned role; JSON output returns:
{
"implemented": false,
"changes": [],
"plannedRole": [
"inspect existing local agent tool config",
"report import candidates for library, defaults, and project config",
"avoid writing files unless a future explicit import/write step is added"
],
"message": "drwn scan is not implemented yet."
}
Planned role
When implemented, drwn scan will:
- inspect existing local agent tool config (Claude, Codex, Cursor, MCP) without mutating any files
- report import candidates for promotion into the library, machine defaults, or current project overlay
- remain non-mutating by construction — the eventual import/promotion step will be a separate explicit command
What it does today
Nothing. The command has no filesystem-mutation imports — read-only by construction (cli/commands/scan.ts). Do not rely on it for any current discovery use case. For inspection today, use:
drwn status— effective harness summarydrwn mcp list— active MCP statedrwn skills list— active skill inventorydrwn doctor— drift, broken links, and project config issues