prism-mcp-server 20.22.1 → 20.22.2

package/README.md

@@ -235,6 +235,18 @@ or by re-enabling after each run.
 <details>
 <summary>Release history (optional)</summary>
 
+## What's New in v20.22.2
+
+**Why it matters:** an agent that followed `session_task_route`'s advice was
+refused on explicit oMLX before anything ran.
+
+- **A complexity hint no longer refuses explicit oMLX.** `session_task_route`
+  recommends a `task_complexity` for every code task. That hint picks Ollama's
+  starting tier, but on explicit oMLX it lowered the model ceiling, so a call
+  that passed the recommended arguments was refused with `omlx_model_ceiling`.
+  Now only an explicit `model_ceiling` below the top tier, or a plan below it,
+  refuses. Thinking on oMLX stays with the server's settings.
+
 ## What's New in v20.22.1
 
 **Why it matters:** on explicit oMLX, a long prompt could keep the device busy

package/dist/tools/prismInferHandler.js

@@ -2206,12 +2206,14 @@ async function runInferAfterInputSafety(args, deps, t0, temperature) {
         // No trusted binding maps an oMLX model to a logical tier, and the
         // configured Flash model outweighs every Ollama tier. Below the top
         // ceiling, from the caller or the plan, refuse before any network,
-        // classification or residency change.
-        if (effectiveCeiling !== "27b") {
+        // classification or residency change. A task_complexity hint only picks
+        // Ollama's starting tier (session_task_route recommends one for every
+        // code task), so it is not a ceiling here; an explicit model_ceiling is.
+        if (clampCeiling(args.model_ceiling ?? "27b", ent.model_ceiling) !== "27b") {
             attempts.push({ tier: "omlx", reason: "omlx_model_ceiling" });
             if (wantReport)
                 return refusedResult("omlx_model_ceiling");
-            throw new Error("prism_infer: omlx_model_ceiling; explicit oMLX needs the top model ceiling");
+            throw new Error("prism_infer: omlx_model_ceiling; explicit oMLX needs the top model ceiling from model_ceiling and the plan");
         }
         if (!Number.isInteger(localMaxTokens) || localMaxTokens <= 0
             || !Number.isFinite(temperature) || temperature < 0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "prism-mcp-server",
-  "version": "20.22.1",
+  "version": "20.22.2",
   "mcpName": "io.github.dcostenco/prism-coder",
   "description": "Persistent session memory for AI coding agents that never leaves your machine — including the on-device model that reasons over it. Restores your prior decisions, open TODOs, and changed files across sessions; adds associative recall of related past work, semantic drift detection, and local inference. Local-first by default. Works with Claude Code, Cursor, and Codex.",
   "module": "index.ts",