Shared memory and context tools for agentic work.
Code Rooms
{
"_about": "X-RAY North Star (MANIFESTO) for the m1nd workspace. Declares the architectural intent that the X-RAY verbs (xray_orient / xray_gate / xray_paint) measure the as-built code against. RATIFIED — the guardrail now blocks EROSION (a dependency that violates a layer rule).",
"ratified": true,
"ratified_on": "2026-06-26",
"layer_order": ["m1nd-core", "m1nd-ingest", "m1nd-mcp", "m1nd-openclaw"],
"forbid": [],
"require_exists": ["mission_verify", "seek", "lock_create", "query_readonly"],
"_layers": {
"m1nd-core": "Foundation — graph, activation, plasticity, snapshot. Depends on no other workspace crate.",
"m1nd-ingest": "Tree-sitter ingestion of source into the graph. May depend on m1nd-core only.",
"m1nd-mcp": "MCP server, handlers, session. May depend on m1nd-core and m1nd-ingest.",
"m1nd-openclaw": "Leaf application. May depend on any lower layer."
},
"_notes": "layer_order alone encodes the no-upward-dependency rule (a crate may depend only on crates at its level or lower); `forbid` is for extra explicit bans beyond layering. `require_exists` lists substrings that must appear in some node's external_id (existence intents -> BLUEPRINT if absent)."
}
