Lumbridge Compute

Governed compute for unified-memory AI hardware — the machines where CPU and
GPU share one pool and there is no separate VRAM allocation to bounce off.
Over-commit that pool and the box thrashes and wedges, SSH and ping included,
before the OOM killer gets a turn.

Compute does not run inference. It supervises the servers that do:

- Admission control against two ceilings: a declared budget, and what the
  machine actually has free. The refusal is the feature.
- A 1 Hz watchdog on MemAvailable that stops the newest model before thrash,
  and defers to a Scene transition rather than racing it.
- Scenes: named sets of models activated as one transactional unit, with
  pre-flight validation and rollback to the previously active Scene on
  failure. Scenes reference model ids, never weight paths or commands, so a
  Scene obtained from elsewhere cannot introduce code.
- Process ownership bound to (boot_id, pid, start_time_ticks, pgid == pid),
  so a reused PID can never be group-killed.
- A protocol-transparent TCP gateway, so clients keep one address while model
  runtimes move behind it.
- An MCP server, so agents drive the node as tools rather than as a CLI.

One binary, no async runtime outside the MCP surface. Apache-2.0.

Generated from the internal monorepo by scripts/publish-compute.sh, which
refuses to publish a tree it cannot prove clean.

evals/smoke.eval.yaml

@@ -0,0 +1,33 @@
+apiVersion: lumbridge/v1
+kind: EvalSuite
+metadata:
+  name: smoke
+  version: 1
+  description: "Fast correctness and serving-health gate for every new model."
+  tags: [smoke, ci]
+defaults:
+  max_tokens: 96
+  temperature: 0.0
+  repeat: 1
+  system: "Follow the requested output format exactly. Do not explain unless asked."
+cases:
+  - id: exact-instruction
+    category: instruction
+    prompt: "Reply with exactly: lumbridge ready"
+    assertions:
+      - type: exact
+        value: "lumbridge ready"
+  - id: arithmetic
+    category: reasoning
+    prompt: "A box has 121 GB. The OS reserves 21 GB and models use 75 GB. Reply with only the remaining number."
+    assertions:
+      - type: exact
+        value: "25"
+  - id: concise-voice
+    category: voice
+    prompt: "In at most twelve words, say that risk limits are operating normally. No markdown."
+    assertions:
+      - type: max_words
+        value: 12
+      - type: not_contains
+        value: "**"