prism-mcp-server 17.0.1 → 17.1.0

package/README.md

@@ -200,7 +200,7 @@ HRR acts as Tier 0 — if confidence is high, FTS5 is skipped entirely. Falls th
 
 Top-1 = correct word is tile #1. MRR = Mean Reciprocal Rank. Zero Top-5 regressions in any scenario. HRR encodes bigrams + trigrams from every spoken phrase; probes take ~0.2ms — safe on every keystroke. All Synalux apps (clinical, AAC, PrismCoach) share HRR via the portal `/api/v1/hrr` endpoint.
 
-**Competitive comparison:**
+**Memory retrieval comparison:**
 
 | System | Retrieval | Offline | Cost | Latency |
 |--------|-----------|---------|------|---------|
@@ -512,16 +512,99 @@ Cascade (14B→32B): **100.0%** · Opus solo: 98.3% · Opus engaged: **0% of req
 
 ## Plans
 
-| Plan | Cloud model | Daily limit | On-device |
-|---|---|---|---|
-| **Free** | — | unlimited local | prism-coder:1.7b (100%) + 8b (100%) + 14b (100%) |
-| **Standard $19/mo** | Claude Sonnet 4 | 200 req | + cloud fallback |
-| **Pro $49/mo** | prism-coder:32b | 2,000 req | + reasoning tier |
-| **Enterprise $99/mo** | prism-coder:32b priority | unlimited | + HIPAA BAA + custom fine-tuning |
-
-All on-device models are **free for every tier** — no subscription needed for local inference. Offline translation (1,261 phrases × 20 languages) included in all plans.
-
-[Subscribe →](https://synalux.ai/pricing)
+| | **Free** | **Standard $19/mo** | **Advanced $49/mo** | **Enterprise $99/mo** |
+|---|---|---|---|---|
+| **Local model ceiling** | up to 4b | up to 14b | up to 32b | up to 32b |
+| **Daily inference limit** | 50 | 200 | 2,000 | 100,000 |
+| **Max output tokens** | 512 | 1,024 | 2,048 | 4,096 |
+| **Cloud fallback** | — | Portal cascade (14b → 32b) | Portal cascade (14b → 32b → Claude Opus) | Priority cascade + Claude Opus |
+| **L3 grounding verifier** | — | ✓ | ✓ | ✓ |
+| **Knowledge search** | limited | unlimited | unlimited | unlimited |
+| **Session memory** | limited | unlimited | unlimited | unlimited |
+| **Analytics dashboard** | — | ✓ | ✓ | ✓ |
+| **HIPAA BAA** | — | — | — | ✓ |
+
+### What free users get
+
+- Local Ollama inference with models up to 4b (prism-coder:1b7 and prism-coder:4b)
+- 50 calls/day, 512 max output tokens per call
+- Local SQLite storage for session memory and knowledge
+- All open-weight models available to pull via `ollama pull`
+
+### What paid users get
+
+- **Higher model ceilings** — Standard unlocks 14b, Advanced/Enterprise unlock 32b
+- **Cloud fallback** — when local Ollama is down or underpowered, inference routes through the Synalux portal cascade (14b → 32b → Claude Opus)
+- **L3 grounding verifier** — evidence-based claim verification that rejects hallucinated outputs
+- **Unlimited knowledge search and session memory** — no caps on stored context
+- **Analytics dashboard** — usage metrics, latency tracking, model performance
+- **Higher daily limits and token caps** — see table above
+
+All on-device models are open-weight and free to run locally via Ollama. The subscription gates cloud features, higher model tiers, and increased limits.
+
+14-day free trial on all paid plans. [Subscribe →](https://synalux.ai/pricing)
+
+### Why Prism MCP
+
+**Pricing — flat-rate, not per-seat:**
+
+| | **Prism MCP** | GitHub Copilot | Cursor | Windsurf | Amazon Q | Tabnine |
+|---|---|---|---|---|---|---|
+| **Individual** | **$19/mo** | $10/mo | $20/mo | $15-20/mo | $19/mo | $39/mo |
+| **Team (5 devs)** | **$49/mo flat** | $95/mo | $200/mo | $200/mo | $95/mo | $295/mo |
+| **Enterprise** | **$99/mo flat** | $195/mo | $1,000/mo | Custom | Custom | Custom |
+
+**Features — full stack vs single-purpose:**
+
+| | **Prism MCP** | GitHub Copilot | Cursor | Windsurf | Amazon Q | Tabnine | Devin |
+|---|---|---|---|---|---|---|---|
+| **Web IDE** | **Synalux Coder** | github.dev | — | — | Console | — | Browser |
+| **VS Code extension** | **Yes** | Yes | N/A (is a fork) | N/A (is a fork) | Yes | Yes | No |
+| **MCP server** | **Native** | No | Partial | No | No | No | No |
+| **Works with Claude Code** | **Yes** | No | N/A | No | No | No | No |
+| **Local inference (Ollama)** | **1.7B–32B fleet** | No | No | No | No | No | No |
+| **Cloud fallback** | **14b→32b→Opus** | Cloud only | Cloud only | Cloud only | Cloud only | Cloud only | Cloud only |
+| **Works offline** | **Yes** | No | No | No | No | No | No |
+| **Open-weight models** | **HuggingFace** | Proprietary | Proprietary | Proprietary | Proprietary | Proprietary | Proprietary |
+| **Persistent memory** | **Cross-session** | No | No | No | No | No | Partial |
+| **Cognitive routing** | **Episodic/semantic/procedural** | No | No | No | No | No | No |
+| **Session drift detection** | **HRR-based** | No | No | No | No | No | No |
+| **Codebase indexing** | **Knowledge ingest (MCP + webhook + REST)** | Partial | Yes | Yes | Yes | Yes | Yes |
+| **L3 grounding verifier** | **Evidence-based** | No | No | No | No | No | No |
+| **Multi-agent hivemind** | **Shared Mind Palace** | No | No | No | No | No | No |
+| **Analytics dashboard** | **Yes** | No | Yes | Yes | Yes | No | Yes |
+| **HIPAA / air-gapped** | **On-prem, no BAA needed** | Requires BAA | No | No | Partial | No | No |
+| **Data stays local** | **Yes** | No | No | No | No | No | No |
+
+**vs local AI tools:**
+
+| | **Prism MCP** | Ollama | LM Studio | Jan.ai | Mem0 | Zep |
+|---|---|---|---|---|---|---|
+| **Local inference** | 1.7B–32B cascade | Any GGUF | Any GGUF | Any GGUF | No | No |
+| **Cloud fallback** | Automatic | No | No | Partial | Cloud only | Cloud only |
+| **Persistent memory** | Cross-session | No | No | No | Yes | Yes |
+| **Knowledge ingestion** | MCP + GitHub webhook + REST | No | No | No | Partial | No |
+| **Cognitive routing** | 3-store (episodic/semantic/procedural) | No | No | No | No | Temporal graph |
+| **Grounding verifier** | L3 evidence-based | No | No | No | No | No |
+| **Drift detection** | HRR-based | No | No | No | No | No |
+| **MCP server** | Native | No | No | No | No | No |
+| **Web IDE** | Synalux Coder | No | No | No | No | No |
+| **VS Code extension** | Yes | No | No | No | No | No |
+| **Analytics** | Dashboard + Datadog | No | No | No | Yes | Yes |
+| **Price** | $0–99/mo flat | Free | Free/$10/user | Free | $249/mo | $99/mo |
+
+**Why developers choose Prism:**
+- **Full IDE experience** — Synalux Coder (web) + VS Code extension + MCP for Claude Code, Cursor, JetBrains
+- **Local-first** — your code and context never leave your machine unless you opt in to cloud
+- **Flat-rate pricing** — $49/mo for your whole team, not $40/seat/mo
+- **Works offline** — airplane, hospital, air-gapped classified environments
+- **Open models** — prism-coder weights are on HuggingFace, not locked behind an API
+- **Memory that persists** — cognitive routing stores episodic, semantic, and procedural memory across sessions
+- **Drift detection** — HRR-based session monitoring catches when your AI agent goes off-track
+- **Grounding verification** — L3 verifier rejects hallucinated outputs before they reach you
+- **Codebase indexing** — knowledge ingestion via MCP tool, GitHub webhooks, or REST API
+- **Multi-agent ready** — Hivemind lets multiple agents share the same Mind Palace with role-scoped context
+- **HIPAA without paperwork** — local inference means no BAA required, PHI never leaves the device
 
 ---
 

package/dist/tools/ledgerHandlers.js

@@ -3,6 +3,7 @@ import * as nodePath from "node:path";
 import * as os from "node:os";
 import { randomUUID } from "node:crypto";
 import { redactSettings, toMarkdown } from "./commonHelpers.js";
+import { scanAndRedactPHI } from "../utils/phiGuard.js";
 import * as fflate from "fflate";
 import { buildVaultDirectory } from "../utils/vaultExporter.js";
 /**
@@ -60,9 +61,11 @@ import { notifyResourceUpdate } from "../server.js";
  * Zero-latency (pure regex, no API calls). Runs on every save.
  */
 export function sanitizeMemoryInput(text) {
-    return text
+    const stripped = text
         .replace(/<\/?(?:system|user_input|instruction|anti_pattern|desired_pattern|assistant|tool_call|prism_memory)[^>]*>/gi, '')
         .trim();
+    // HIPAA: redact PHI before storage — SSN, DOB, MRN, patient names, etc.
+    return scanAndRedactPHI(stripped).redacted;
 }
 /** Sanitize each string in an array (for decisions[], todos[], etc.) */
 function sanitizeArray(arr) {
@@ -853,17 +856,29 @@ export async function sessionLoadContextHandler(args) {
             .fetchSkillContent(missing).catch(() => ({}));
         debugLog(`[session_load_context] Synalux skill content fetched: ${Object.keys(synaluxContent).join(", ") || "none"}`);
     }
+    const SKILL_BLOCK_CAP = 30_000;
+    const skippedSkills = [];
     for (const skillName of skillsToLoad) {
         if (loadedSkills.includes(skillName))
             continue;
-        // Synalux (paid) → platform fallback skill:<name> (free/offline). Never user_skill:.
+        if (skillBlock.length >= SKILL_BLOCK_CAP) {
+            skippedSkills.push(skillName);
+            debugLog(`[session_load_context] Skill "${skillName}" skipped — block cap ${SKILL_BLOCK_CAP} reached`);
+            continue;
+        }
         const content = synaluxContent[skillName] || await getSetting(`skill:${skillName}`, "");
         if (content && content.trim()) {
+            const trimmed = content.trim();
+            if (skillBlock.length + trimmed.length > SKILL_BLOCK_CAP && loadedSkills.length > 0) {
+                skippedSkills.push(skillName);
+                debugLog(`[session_load_context] Skill "${skillName}" skipped — would exceed cap (${skillBlock.length}+${trimmed.length} > ${SKILL_BLOCK_CAP})`);
+                continue;
+            }
             const source = synaluxContent[skillName] ? "synalux" : "local-platform";
-            skillBlock += `\n\n[📜 SKILL: ${skillName}]\n${content.trim()}`;
+            skillBlock += `\n\n[📜 SKILL: ${skillName}]\n${trimmed}`;
             loadedSkills.push(skillName);
             skillLoaded = true;
-            debugLog(`[session_load_context] Skill "${skillName}" loaded (${source}) for project="${project}"`);
+            debugLog(`[session_load_context] Skill "${skillName}" loaded (${source}) for project="${project}" [${skillBlock.length}/${SKILL_BLOCK_CAP} chars]`);
         }
     }
     // ─── User-Local Skills ──────────────────────────────────────
@@ -876,10 +891,15 @@ export async function sessionLoadContextHandler(args) {
         for (const [k, v] of Object.entries(allSettings)) {
             if (!k.startsWith(prefix) || !v)
                 continue;
+            if (skillBlock.length >= SKILL_BLOCK_CAP)
+                break;
             const skillName = k.replace(prefix, "");
             if (loadedSkills.includes(skillName))
                 continue;
-            skillBlock += `\n\n[📜 USER SKILL: ${skillName}]\n${v.trim()}`;
+            const trimmed = v.trim();
+            if (skillBlock.length + trimmed.length > SKILL_BLOCK_CAP && loadedSkills.length > 0)
+                continue;
+            skillBlock += `\n\n[📜 USER SKILL: ${skillName}]\n${trimmed}`;
             loadedSkills.push(skillName);
             skillLoaded = true;
             debugLog(`[session_load_context] User-local skill "${skillName}" loaded`);
@@ -888,22 +908,32 @@ export async function sessionLoadContextHandler(args) {
     // ─── Memory-Based Skill Discovery ──────────────────────────
     // If recent handoff/ledger mentions a platform skill name, auto-load it.
     // Only scans platform skill: keys — user_skill: discovery is not automatic.
-    if (formattedContext.length > 0) {
+    if (formattedContext.length > 0 && skillBlock.length < SKILL_BLOCK_CAP) {
         const contextText = formattedContext.toLowerCase();
         const allSkillKeys = await storage.getAllSettings?.() || {};
         for (const [k, v] of Object.entries(allSkillKeys)) {
             if (!k.startsWith("skill:") || !v)
                 continue;
+            if (skillBlock.length >= SKILL_BLOCK_CAP)
+                break;
             const skillName = k.replace("skill:", "");
             if (loadedSkills.includes(skillName))
                 continue;
             if (contextText.includes(skillName.replace(/-/g, " ")) || contextText.includes(skillName)) {
-                skillBlock += `\n\n[📜 CONTEXT SKILL: ${skillName}]\n${v}`;
+                const trimmed = v.trim();
+                if (skillBlock.length + trimmed.length > SKILL_BLOCK_CAP && loadedSkills.length > 0) {
+                    skippedSkills.push(skillName);
+                    continue;
+                }
+                skillBlock += `\n\n[📜 CONTEXT SKILL: ${skillName}]\n${trimmed}`;
                 loadedSkills.push(skillName);
                 debugLog(`[session_load_context] Context-triggered skill "${skillName}"`);
             }
         }
     }
+    if (skippedSkills.length > 0) {
+        skillBlock += `\n\n[⏭️ ${skippedSkills.length} skills skipped (cap ${SKILL_BLOCK_CAP} chars): ${skippedSkills.join(", ")}]`;
+    }
     // ─── Agent Greeting Block ────────────────────────────────────
     // Shows agent identity (name + role) and skill status after briefing.
     let greetingBlock = "";

package/dist/tools/prismInferHandler.js

@@ -25,6 +25,8 @@ import { getAvailableMemoryBytes } from "../utils/availableMemory.js";
 import { PRISM_SYNALUX_BASE_URL, PRISM_LOCAL_LLM_URL, } from "../config.js";
 import { debugLog } from "../utils/logger.js";
 import { verifyGrounding } from "../utils/groundingVerifier.js";
+import { getEntitlements, clampCeiling } from "../utils/entitlements.js";
+import { ddLog } from "../utils/ddLogger.js";
 // ─── Tool Definition ────────────────────────────────────────────
 export const PRISM_INFER_TOOL = {
     name: "prism_infer",
@@ -273,12 +275,47 @@ async function callSynaluxInference(prompt, maxTokens, timeoutMs) {
 }
 export async function runInfer(args, deps) {
     const t0 = Date.now();
-    const maxTokens = Math.min(args.max_tokens ?? 1024, 8192);
     const temperature = args.temperature ?? 0;
-    const allowCloud = args.cloud_fallback === true;
+    // ── Entitlement enforcement ──────────────────────────────────
+    // Fetch user's plan limits (cached 1hr). Free users without auth
+    // get 4b ceiling, 50 calls/day, 512 max tokens.
+    const ent = deps.entitlements ?? await getEntitlements();
+    // Clamp model ceiling to what the plan allows
+    const effectiveCeiling = clampCeiling(args.model_ceiling, ent.model_ceiling);
+    // Clamp max_tokens to plan limit
+    const maxTokens = Math.min(args.max_tokens ?? 1024, ent.max_tokens, 8192);
+    // Cloud fallback only for paid plans
+    const allowCloud = args.cloud_fallback === true && ent.features.cloud_fallback;
+    // Verification only for paid plans (free users skip L3 grounding)
+    const canVerify = ent.features.grounding_verifier;
     const freeBytes = deps.freemem();
     const ramFreeMb = Math.round(freeBytes / (1024 * 1024));
     const attempts = [];
+    // Strip verification args if plan lacks grounding_verifier
+    const gatedArgs = canVerify ? args : { ...args, verify: false, evidence: undefined };
+    debugLog(`[prism_infer] plan=${ent.plan} ceiling=${effectiveCeiling} max_tokens=${maxTokens} cloud=${allowCloud} verify=${canVerify}`);
+    // Log tier enforcement to Datadog for monetization visibility
+    const ceilingClamped = effectiveCeiling !== (args.model_ceiling ?? ent.model_ceiling);
+    const tokensClamped = maxTokens < (args.max_tokens ?? 1024);
+    const cloudBlocked = args.cloud_fallback === true && !allowCloud;
+    const verifierBlocked = (args.verify === true || (args.evidence?.length ?? 0) > 0) && !canVerify;
+    if (ceilingClamped || tokensClamped || cloudBlocked || verifierBlocked) {
+        ddLog("info", "prism_infer.tier_enforcement", {
+            plan: ent.plan,
+            requested_ceiling: args.model_ceiling,
+            effective_ceiling: effectiveCeiling,
+            ceiling_clamped: ceilingClamped,
+            requested_tokens: args.max_tokens,
+            effective_tokens: maxTokens,
+            tokens_clamped: tokensClamped,
+            cloud_requested: args.cloud_fallback,
+            cloud_allowed: allowCloud,
+            cloud_blocked: cloudBlocked,
+            verify_requested: args.verify,
+            verify_allowed: canVerify,
+            verify_blocked: verifierBlocked,
+        });
+    }
     // Discover which tags Ollama actually has + which are already warm.
     // Already-loaded models don't need RAM headroom — they're reusing
     // memory Ollama allocated previously.
@@ -292,8 +329,8 @@ export async function runInfer(args, deps) {
     // so the caller can see exactly why each tier was bypassed.
     if (installed) {
         // Find start index from ceiling — if no ceiling, start at the top (32B).
-        const ceilStart = args.model_ceiling
-            ? Math.max(0, MODEL_TIERS.findIndex(t => t.tag.endsWith(args.model_ceiling) || t.tag === args.model_ceiling))
+        const ceilStart = effectiveCeiling
+            ? Math.max(0, MODEL_TIERS.findIndex(t => t.tag.endsWith(effectiveCeiling) || t.tag === effectiveCeiling))
             : 0;
         let anyViable = false;
         for (let i = ceilStart; i < MODEL_TIERS.length; i++) {
@@ -318,13 +355,14 @@ export async function runInfer(args, deps) {
             const timeout = args.timeout_ms ?? DEFAULT_TIMEOUTS[tier.tag] ?? 60_000;
             const result = await deps.callLocal(deps.ollamaUrl, ollamaName, args.prompt, args.system, maxTokens, temperature, timeout);
             if (result.ok) {
-                return await applyVerification(result.text, args, deps, {
+                return await applyVerification(result.text, gatedArgs, deps, {
                     backend: `ollama-${tier.tag.replace("prism-coder:", "")}`,
                     model_picked: tier.tag,
                     ram_free_mb: ramFreeMb,
                     latency_ms: Date.now() - t0,
                     used_cloud: false,
                     attempts,
+                    plan: ent.plan,
                 });
             }
             attempts.push({ tier: tier.tag, reason: result.reason });
@@ -340,13 +378,14 @@ export async function runInfer(args, deps) {
         const cloudTimeout = args.timeout_ms ?? 90_000;
         const cloud = await deps.callCloud(args.prompt, maxTokens, cloudTimeout);
         if (cloud.ok && cloud.output) {
-            return await applyVerification(cloud.output, args, deps, {
+            return await applyVerification(cloud.output, gatedArgs, deps, {
                 backend: cloud.backend ?? "synalux",
                 model_picked: null,
                 ram_free_mb: ramFreeMb,
                 latency_ms: Date.now() - t0,
                 used_cloud: true,
                 attempts,
+                plan: ent.plan,
             });
         }
         attempts.push({ tier: "synalux", reason: cloud.reason ?? "unknown" });
@@ -408,6 +447,7 @@ export async function prismInferHandler(args) {
         debugLog(`[prism_infer] backend=${result.backend} model=${result.model_picked} latency=${result.latency_ms}ms free=${result.ram_free_mb}MB`);
         const header = `[prism_infer] backend=${result.backend}` +
             ` model=${result.model_picked ?? "n/a"}` +
+            ` plan=${result.plan ?? "unknown"}` +
             ` free_ram=${result.ram_free_mb}MB` +
             ` latency=${result.latency_ms}ms` +
             ` used_cloud=${result.used_cloud}` +

package/dist/tools/skillRouting.js

@@ -81,6 +81,45 @@ export async function resolveSkillsForProject(project) {
         user_local: table.user_local ?? OFFLINE_FALLBACK.user_local,
     };
 }
+/**
+ * Resolve skills based on user prompt keywords. Matches prompt text
+ * against the routing table's prompt_keywords regex patterns.
+ * Returns deduplicated skill names (excluding any already in baseSkills).
+ */
+export async function resolveSkillsForPrompt(prompt, baseSkills = []) {
+    const now = Date.now();
+    if (!cached || now - cached.fetchedAt > CACHE_TTL_MS) {
+        if (!inflight) {
+            inflight = fetchOnce().then((table) => {
+                cached = { table, fetchedAt: Date.now() };
+                return table;
+            }).finally(() => { inflight = null; });
+        }
+        await inflight;
+    }
+    const table = cached.table;
+    if (!table.prompt_keywords)
+        return [];
+    const existing = new Set(baseSkills);
+    const matched = [];
+    for (const [pattern, skills] of Object.entries(table.prompt_keywords)) {
+        try {
+            const re = new RegExp(pattern, 'i');
+            if (re.test(prompt)) {
+                for (const s of skills) {
+                    if (!existing.has(s)) {
+                        existing.add(s);
+                        matched.push(s);
+                    }
+                }
+            }
+        }
+        catch {
+            // Invalid regex in routing table — skip silently
+        }
+    }
+    return matched;
+}
 /** Force a re-fetch on the next call. Exposed for tests + admin tooling. */
 export function _invalidateRoutingCache() {
     cached = null;

package/dist/utils/entitlements.js

@@ -0,0 +1,136 @@
+/**
+ * Prism Entitlements — Plan-Based Feature & Model Gating
+ * ═══════════════════════════════════════════════════════════
+ * Fetches the user's plan entitlements from the Synalux portal
+ * and caches them locally. Used by prism_infer and other tools
+ * to enforce model ceiling, max_tokens, and feature gates.
+ *
+ * Unauthenticated users (no SYNALUX_API_KEY) get free-tier defaults.
+ * Authenticated users get their plan from the portal (1-hour cache).
+ */
+import { getSynaluxJwt } from "./synaluxJwt.js";
+import { PRISM_SYNALUX_BASE_URL, SYNALUX_CONFIGURED } from "../config.js";
+import { debugLog } from "./logger.js";
+// ── Free-tier defaults (no auth) ──────────────────────────────────
+export const FREE_ENTITLEMENTS = {
+    plan: "free",
+    model_ceiling: "4b",
+    daily_infer_limit: 50,
+    max_tokens: 512,
+    features: {
+        cloud_fallback: false,
+        grounding_verifier: false,
+        knowledge_search_unlimited: false,
+        session_memory_unlimited: false,
+        analytics_dashboard: false,
+    },
+    upgrade_url: "https://synalux.ai/pricing",
+};
+// ── Cache ─────────────────────────────────────────────────────────
+const CACHE_TTL_MS = 60 * 60 * 1000; // 1 hour
+let cache = null;
+let inFlight = null;
+// ── Model tier ordering for ceiling enforcement ───────────────────
+const TIER_ORDER = ["1b7", "4b", "8b", "14b", "32b"];
+/**
+ * Returns true if `requested` exceeds `ceiling`.
+ * e.g. ceilingExceeded("14b", "4b") → true (14b > 4b ceiling)
+ */
+export function ceilingExceeded(requested, ceiling) {
+    const reqIdx = TIER_ORDER.indexOf(requested);
+    const ceilIdx = TIER_ORDER.indexOf(ceiling);
+    if (reqIdx === -1 || ceilIdx === -1)
+        return false;
+    return reqIdx > ceilIdx;
+}
+/**
+ * Clamp a model ceiling string to the plan's maximum.
+ * Returns the lower of the two ceilings.
+ */
+export function clampCeiling(requested, planCeiling) {
+    if (!requested)
+        return planCeiling;
+    const reqIdx = TIER_ORDER.indexOf(requested);
+    const planIdx = TIER_ORDER.indexOf(planCeiling);
+    if (reqIdx === -1)
+        return planCeiling;
+    if (planIdx === -1)
+        return requested;
+    return TIER_ORDER[Math.min(reqIdx, planIdx)];
+}
+// ── Fetch ─────────────────────────────────────────────────────────
+async function fetchEntitlements() {
+    if (!SYNALUX_CONFIGURED || !PRISM_SYNALUX_BASE_URL) {
+        debugLog("[entitlements] no Synalux auth configured — free tier");
+        return FREE_ENTITLEMENTS;
+    }
+    const jwt = await getSynaluxJwt();
+    if (!jwt) {
+        debugLog("[entitlements] JWT exchange failed — free tier fallback");
+        return FREE_ENTITLEMENTS;
+    }
+    try {
+        const url = `${PRISM_SYNALUX_BASE_URL}/api/v1/prism/entitlements`;
+        const res = await fetch(url, {
+            method: "GET",
+            headers: { Authorization: `Bearer ${jwt}` },
+            signal: AbortSignal.timeout(10_000),
+            redirect: "error",
+        });
+        if (!res.ok) {
+            debugLog(`[entitlements] portal HTTP ${res.status} — free tier fallback`);
+            return FREE_ENTITLEMENTS;
+        }
+        const data = (await res.json());
+        if (!data.plan || !data.model_ceiling) {
+            debugLog("[entitlements] malformed response — free tier fallback");
+            return FREE_ENTITLEMENTS;
+        }
+        debugLog(`[entitlements] plan=${data.plan} ceiling=${data.model_ceiling} ` +
+            `daily=${data.daily_infer_limit} max_tokens=${data.max_tokens}`);
+        return data;
+    }
+    catch (err) {
+        debugLog(`[entitlements] fetch error: ${err instanceof Error ? err.message : String(err)} — free tier fallback`);
+        return FREE_ENTITLEMENTS;
+    }
+}
+// ── Public API ────────────────────────────────────────────────────
+/**
+ * Get the current user's entitlements (cached for 1 hour).
+ * Concurrent callers share a single in-flight fetch.
+ */
+export async function getEntitlements() {
+    const now = Date.now();
+    if (cache && cache.expiresAt > now) {
+        return cache.entitlements;
+    }
+    if (inFlight)
+        return inFlight;
+    inFlight = (async () => {
+        try {
+            const ent = await fetchEntitlements();
+            cache = { entitlements: ent, expiresAt: Date.now() + CACHE_TTL_MS };
+            return ent;
+        }
+        finally {
+            inFlight = null;
+        }
+    })();
+    return inFlight;
+}
+/**
+ * Force cache invalidation (e.g. after plan upgrade).
+ */
+export function invalidateEntitlements() {
+    cache = null;
+}
+/** Test-only: reset all state. */
+export function _resetEntitlementsForTest() {
+    cache = null;
+    inFlight = null;
+}
+/** Test-only: inject a cached entitlement. */
+export function _setCacheForTest(ent, ttlMs = CACHE_TTL_MS) {
+    cache = { entitlements: ent, expiresAt: Date.now() + ttlMs };
+}

package/dist/utils/phiGuard.js

@@ -0,0 +1,88 @@
+/**
+ * PHI Guard — detect and redact Protected Health Information before storage/logging.
+ *
+ * HIPAA §164.502: PHI must not be disclosed except as permitted.
+ * This module scans text for common PHI patterns (SSN, DOB, MRN, phone,
+ * email, patient names in clinical context) and redacts them.
+ *
+ * Detection events are logged to stderr (picked up by DD agent) with
+ * the pattern type and character position — never the actual PHI value.
+ *
+ * Usage:
+ *   import { scanAndRedactPHI, hasPHI } from './phiGuard.js';
+ *   const { redacted, detections } = scanAndRedactPHI(userText);
+ *   // `redacted` is safe to store/log; `detections` lists what was found
+ */
+// Patterns ordered by specificity (most specific first)
+const PHI_PATTERNS = [
+    // SSN: 123-45-6789 or 123456789
+    { name: 'SSN', regex: /\b\d{3}-\d{2}-\d{4}\b/g, replacement: '[SSN-REDACTED]' },
+    { name: 'SSN', regex: /\b\d{9}\b(?=\s|$|[,.])/g, replacement: '[SSN-REDACTED]' },
+    // Date of birth patterns: DOB: 01/15/1990, born 1990-01-15, birthday 01/15/90
+    { name: 'DOB', regex: /\b(?:dob|date\s*of\s*birth|born|birthday)\s*[:=]?\s*\d{1,2}[/\-]\d{1,2}[/\-]\d{2,4}\b/gi, replacement: '[DOB-REDACTED]' },
+    { name: 'DOB', regex: /\b(?:dob|date\s*of\s*birth|born|birthday)\s*[:=]?\s*\d{4}[/\-]\d{1,2}[/\-]\d{1,2}\b/gi, replacement: '[DOB-REDACTED]' },
+    // Medical Record Number: MRN: 12345678, MRN#12345
+    { name: 'MRN', regex: /\b(?:mrn|medical\s*record)\s*[#:=]?\s*\d{4,12}\b/gi, replacement: '[MRN-REDACTED]' },
+    // US Phone: (301) 433-1943, 301-433-1943, +1-301-433-1943
+    { name: 'PHONE', regex: /\b(?:\+?1[-.]?)?\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}\b/g, replacement: '[PHONE-REDACTED]' },
+    // Email in clinical context: patient email, client email
+    { name: 'EMAIL', regex: /\b(?:patient|client|parent|caregiver)\s*(?:email|e-mail)\s*[:=]?\s*[\w.+-]+@[\w.-]+\.\w{2,}\b/gi, replacement: '[EMAIL-REDACTED]' },
+    // Patient/client name patterns: "Patient: John Doe", "Client Name: Jane Smith"
+    { name: 'PATIENT_NAME', regex: /\b(?:patient|client)\s*(?:name)?\s*[:=]\s*[A-Z][a-z]+\s+[A-Z][a-z]+/gi, replacement: '[NAME-REDACTED]' },
+    // Insurance ID: Ins#, Policy#, Member ID
+    { name: 'INSURANCE_ID', regex: /\b(?:ins(?:urance)?|policy|member)\s*(?:id|#|number)\s*[:=]?\s*[A-Z0-9]{6,20}\b/gi, replacement: '[INSURANCE-REDACTED]' },
+    // Diagnosis codes in patient context: "diagnosed with F84.0", "ICD: F32.1"
+    { name: 'DIAGNOSIS', regex: /\b(?:diagnos\w*|icd|dx)\s*(?:[:=]|with)?\s*[A-Z]\d{2}(?:\.\d{1,2})?\b/gi, replacement: '[DX-REDACTED]' },
+];
+/**
+ * Scan text for PHI patterns and return redacted version + detection list.
+ * Never logs or stores the actual PHI values — only type + position.
+ */
+export function scanAndRedactPHI(text) {
+    if (typeof text !== 'string' || !text) {
+        return { redacted: text || '', detections: [], hasPHI: false };
+    }
+    const detections = [];
+    let redacted = text;
+    for (const { name, regex, replacement } of PHI_PATTERNS) {
+        // Reset regex state for global patterns
+        regex.lastIndex = 0;
+        let match;
+        while ((match = regex.exec(text)) !== null) {
+            detections.push({
+                type: name,
+                position: match.index,
+                length: match[0].length,
+            });
+        }
+        redacted = redacted.replace(regex, replacement);
+    }
+    if (detections.length > 0) {
+        // Log detection event — type + count only, NEVER the actual value
+        const summary = detections.reduce((acc, d) => {
+            acc[d.type] = (acc[d.type] || 0) + 1;
+            return acc;
+        }, {});
+        const summaryStr = Object.entries(summary).map(([k, v]) => `${k}=${v}`).join(' ');
+        console.error(`[PHI-GUARD] Detected and redacted PHI: ${summaryStr}`);
+    }
+    return {
+        redacted,
+        detections,
+        hasPHI: detections.length > 0,
+    };
+}
+/**
+ * Quick check — does the text contain PHI patterns?
+ * Faster than full redaction when you only need a boolean.
+ */
+export function hasPHI(text) {
+    if (typeof text !== 'string' || !text)
+        return false;
+    for (const { regex } of PHI_PATTERNS) {
+        regex.lastIndex = 0;
+        if (regex.test(text))
+            return true;
+    }
+    return false;
+}

package/package.json

@@ -1,8 +1,8 @@
 {
   "name": "prism-mcp-server",
-  "version": "17.0.1",
+  "version": "17.1.0",
   "mcpName": "io.github.dcostenco/prism-coder",
-  "description": "Prism Coder — Cognitive memory + tool-calling intelligence for AI agents. Mind Palace persistent memory (BFCL Gold Certified, 100% Tool-Call Accuracy, 54 Agent Skills, Zero-Search HDC/HRR retrieval, HRR Semantic Drift Detection across BCBA/Coding/AAC domains, HIPAA-hardened local-first storage, SLERP-optimized GRPO alignment) plus the prism-coder:7b / 14b open-weights LLM fleet.",
+  "description": "Prism Coder — Cognitive memory + tool-calling intelligence for AI agents. Mind Palace persistent memory (BFCL Gold Certified, 100% Tool-Call Accuracy, 114 Agent Skills, Zero-Search HDC/HRR retrieval, HRR Semantic Drift Detection across BCBA/Coding/AAC domains, HIPAA-hardened local-first storage, SLERP-optimized GRPO alignment) plus the prism-coder:7b / 14b open-weights LLM fleet.",
   "module": "index.ts",
   "type": "module",
   "main": "dist/server.js",