chat-recall 0.6.4 → 0.6.5

package/dist/mcp-relay.js

@@ -118,7 +118,7 @@ function daemonEnabled() {
 }
 
 // src/mcp-relay.ts
-var VERSION = true ? "0.6.4" : "0.0.0";
+var VERSION = true ? "0.6.5" : "0.0.0";
 var SERVER_URL = new URL("./mcp.js", import.meta.url);
 var SERVER_PATH = fileURLToPath(SERVER_URL);
 var START_TIMEOUT_MS = 15e3;

package/dist/mcp.js

@@ -15259,12 +15259,12 @@ function extractEntities(text, context = {}) {
         if (chosen && looksLikeStrictTool(chosen) && looksLikeStrictTool(rejected)) {
           const c = canonTool(chosen), r = canonTool(rejected);
           if (projectName)
-            addTriple(projectName, "chose", c, 0.8);
-          addTriple(projectName || "project", "rejected", r, 0.7);
-          addTriple(c, "chosen_over", r, 0.8);
+            addTriple(projectName, "chose", c, DECISION_GUESS_CONFIDENCE);
+          addTriple(projectName || "project", "rejected", r, DECISION_GUESS_CONFIDENCE);
+          addTriple(c, "chosen_over", r, DECISION_GUESS_CONFIDENCE);
         }
       } else if (chosen && looksLikeDecisionObject(chosen) && projectName) {
-        addTriple(projectName, "chose", canonTool(chosen), 0.8);
+        addTriple(projectName, "chose", canonTool(chosen), DECISION_GUESS_CONFIDENCE);
       }
     }
   }
@@ -15301,7 +15301,7 @@ function extractEntities(text, context = {}) {
   }
   return triples;
 }
-var KNOWN_TOOLS, AMBIGUOUS_TOOLS, toolContextRe, TOOL_PATTERNS, DECISION_PATTERNS2, DECISION_OBJECT_STOPWORDS, KNOWN_TOOL_NAMES, CANONICAL_TOOL, canonTool, PROJECT_PATH_PATTERN, PEOPLE_PATTERNS, NOT_PEOPLE;
+var KNOWN_TOOLS, AMBIGUOUS_TOOLS, toolContextRe, TOOL_PATTERNS, DECISION_PATTERNS2, DECISION_OBJECT_STOPWORDS, KNOWN_TOOL_NAMES, CANONICAL_TOOL, canonTool, DECISION_GUESS_CONFIDENCE, PROJECT_PATH_PATTERN, PEOPLE_PATTERNS, NOT_PEOPLE;
 var init_entity_extractor = __esm({
   "../engine/dist/core/entity-extractor.js"() {
     "use strict";
@@ -15522,6 +15522,7 @@ var init_entity_extractor = __esm({
       "nextjs": "next.js"
     };
     canonTool = (n) => CANONICAL_TOOL[n.toLowerCase()] ?? n;
+    DECISION_GUESS_CONFIDENCE = 0.6;
     PROJECT_PATH_PATTERN = /\/(?:home|Users)\/\w+\/(?:code|projects|work|dev|src)\/(?:\w+\/)?([a-zA-Z][a-zA-Z0-9_-]{1,40})/g;
     PEOPLE_PATTERNS = [
       /\b(?:my name is|I'm|I am)\s+([A-Z][a-z]{1,15}(?:\s+[A-Z][a-z]{1,15})?)/g,
@@ -19727,7 +19728,7 @@ var init_http = __esm({
       CLI_VERSION = JSON.parse(readFileSync35(new URL("../package.json", import.meta.url), "utf-8")).version || CLI_VERSION;
     } catch {
       try {
-        if (true) CLI_VERSION = "0.6.4";
+        if (true) CLI_VERSION = "0.6.5";
       } catch {
       }
     }
@@ -20247,7 +20248,7 @@ var init_telemetry = __esm({
     init_sync_client();
     cliVersion2 = () => {
       try {
-        return true ? "0.6.4" : "dev";
+        return true ? "0.6.5" : "dev";
       } catch {
         return "dev";
       }
@@ -22157,7 +22158,7 @@ var init_sync_client = __esm({
       ownVersion = JSON.parse(readFileSync37(join48(dir, "../package.json"), "utf-8")).version || ownVersion;
     } catch {
       try {
-        if (true) ownVersion = "0.6.4";
+        if (true) ownVersion = "0.6.5";
       } catch {
       }
     }
@@ -23008,6 +23009,118 @@ function resumeCommandFor(sessionId, tool) {
 
 // ../engine/dist/mcp/tools.js
 init_project_resolver();
+
+// ../engine/dist/core/decision-areas.js
+var DECISION_AREAS = [
+  "auth",
+  "database",
+  "payments",
+  "api",
+  "frontend",
+  "deploy",
+  "testing",
+  "observability",
+  "security",
+  "pricing",
+  "licensing",
+  "positioning"
+];
+var AREA_SET = new Set(DECISION_AREAS);
+var AREA_ALIASES = {
+  // auth
+  authentication: "auth",
+  authn: "auth",
+  authz: "auth",
+  authorisation: "auth",
+  authorization: "auth",
+  login: "auth",
+  "sign-in": "auth",
+  signin: "auth",
+  sso: "auth",
+  identity: "auth",
+  // database
+  db: "database",
+  datastore: "database",
+  storage: "database",
+  persistence: "database",
+  orm: "database",
+  migrations: "database",
+  // payments
+  billing: "payments",
+  payment: "payments",
+  checkout: "payments",
+  subscriptions: "payments",
+  // api
+  backend: "api",
+  server: "api",
+  rpc: "api",
+  // frontend
+  ui: "frontend",
+  client: "frontend",
+  web: "frontend",
+  styling: "frontend",
+  css: "frontend",
+  // deploy
+  deployment: "deploy",
+  infra: "deploy",
+  infrastructure: "deploy",
+  hosting: "deploy",
+  ci: "deploy",
+  cd: "deploy",
+  "ci-cd": "deploy",
+  release: "deploy",
+  // testing
+  tests: "testing",
+  test: "testing",
+  qa: "testing",
+  e2e: "testing",
+  // observability
+  monitoring: "observability",
+  logging: "observability",
+  metrics: "observability",
+  alerting: "observability",
+  tracing: "observability",
+  // security
+  secrets: "security",
+  compliance: "security",
+  privacy: "security",
+  // pricing
+  price: "pricing",
+  plans: "pricing",
+  tiers: "pricing",
+  // licensing
+  licence: "licensing",
+  license: "licensing",
+  // positioning
+  marketing: "positioning",
+  branding: "positioning",
+  messaging: "positioning"
+};
+function canonArea(raw) {
+  if (!raw)
+    return null;
+  const lowered = raw.trim().toLowerCase();
+  if (!lowered)
+    return null;
+  const slug = lowered.replace(/[\s_/]+/g, "-").replace(/[^a-z0-9-]/g, "").replace(/-{2,}/g, "-").replace(/^-|-$/g, "");
+  if (!slug)
+    return null;
+  if (AREA_SET.has(slug))
+    return slug;
+  if (AREA_ALIASES[slug])
+    return AREA_ALIASES[slug];
+  return slug.slice(0, 40);
+}
+function isKnownArea(key2) {
+  return !!key2 && AREA_SET.has(key2);
+}
+var ACCOUNT_SCOPE = "*";
+function decisionSubject(project, area) {
+  const p = (project || "").trim() || ACCOUNT_SCOPE;
+  return `${p}:${area}`;
+}
+
+// ../engine/dist/mcp/tools.js
 import { config } from "dotenv";
 
 // ../engine/dist/mcp/diagnostics.js
@@ -23856,10 +23969,20 @@ var RecallDecisionRecordSchema = z.object({
   subject: z.string().describe('What the decision is about (e.g., "chat-recall", "auth strategy")'),
   decision: z.string().describe('The decision itself in plain words (e.g., "use Postgres full-text search as the default backend")'),
   reason: z.string().optional().describe("Why this was decided \u2014 short rationale"),
+  // The key a reversal supersedes on. Without it the subject is free text, so
+  // "auth", "authentication" and "auth setup" are three separate facts and a
+  // later decision never closes an earlier one — see core/decision-areas.ts.
+  area: z.string().optional().describe(`The area this decides, so a later decision about the SAME area supersedes it. Use one of: ${DECISION_AREAS.join(", ")} \u2014 or any short slug for an area not listed. Without it the decision is recorded but nothing it contradicts is closed.`),
+  project: z.string().optional().describe("Project this decision applies to. Omit for an account-wide decision that applies everywhere."),
   importance: z.number().min(1).max(5).optional().default(4).describe("1\u20135; the classifier surfaces 4+ in wake-up context"),
   session_id: z.string().optional().describe("Session this decision was made in (for traceability)"),
   agent_name: z.string().optional().default("agent").describe("Who recorded the decision (for diary linkage)")
 });
+var RecallDecisionsSchema = z.object({
+  area: z.string().optional().describe("Only this area"),
+  project: z.string().optional().describe("Resolve for this project"),
+  include_candidates: z.boolean().optional().default(false)
+});
 var RecallAnalyticsSummarySchema = z.object({
   // No required inputs — returns the same summary the dashboard renders.
   view: z.enum(["summary", "patterns"]).optional().default("summary").describe("summary = spend/activity rollup; patterns = behavioural patterns across sessions")
@@ -24008,6 +24131,7 @@ var LEAN_TOOLS = /* @__PURE__ */ new Set([
   "recall_kg_query",
   "recall_kg_add",
   "recall_decision_record",
+  "recall_decisions",
   "recall_diary_write",
   "recall_diary_read",
   // Health, and the one that pays out on day one.
@@ -24904,13 +25028,19 @@ I asking yesterday?" or "what did I tell the agent in this session?".`,
         name: "recall_decision_record",
         description: `Record an explicit decision so it shows up in wake-up context. Writes a KG triple
 (subject \u2192 decided \u2192 decision) plus a diary entry. Use this when you and the user agree on
-something non-obvious that future sessions should remember.`,
+something non-obvious that future sessions should remember.
+
+PASS \`area\`. It is the key a reversal supersedes on: recording a new \`auth\` decision closes the
+previous \`auth\` one and leaves every other area alone. Without it the decision is stored but
+nothing it contradicts is closed, and both answers stay live forever.`,
         inputSchema: {
           type: "object",
           properties: {
             subject: { type: "string", description: "What the decision is about" },
             decision: { type: "string", description: "The decision itself" },
             reason: { type: "string", description: "Why this was decided" },
+            area: { type: "string", description: `The area this decides, so a later decision on the same area supersedes it. Standard areas: ${DECISION_AREAS.join(", ")}. Any short slug works for an area not listed.` },
+            project: { type: "string", description: "Project this applies to. Omit for an account-wide decision that applies everywhere." },
             importance: { type: "number", minimum: 1, maximum: 5, default: 4, description: "1-5. Only 4 and 5 surface in the wake-up bundle, so reserve those for decisions a future session must not miss." },
             session_id: { type: "string", description: "Session this decision came out of. Pass your own current session id so the decision can be traced back to the conversation that made it." },
             agent_name: { type: "string", default: "agent", description: "Which agent is recording this, so diary entries stay separable per agent." }
@@ -24918,6 +25048,25 @@ something non-obvious that future sessions should remember.`,
           required: ["subject", "decision"]
         }
       },
+      {
+        name: "recall_decisions",
+        description: `What this codebase has already settled \u2014 read it BEFORE choosing a library, a
+datastore, an auth method or anything else in a decided area, and you will not rebuild what was
+already ruled out.
+
+Returns one effective value per area with the cascade already resolved (project beats account;
+personal preferences fill gaps and never override a team decision), what each value replaced and
+when, and the session it came from. Also returns \`gaps\` \u2014 areas nobody has decided, where you are
+free to choose, and where recording the choice afterwards is worth doing.`,
+        inputSchema: {
+          type: "object",
+          properties: {
+            area: { type: "string", description: `Only this area. Standard areas: ${DECISION_AREAS.join(", ")}.` },
+            project: { type: "string", description: "Resolve for this project, so its overrides win over account-wide decisions. Omit for the account-wide view." },
+            include_candidates: { type: "boolean", default: false, description: "Also return unconfirmed guesses the indexer extracted. These are NOT decisions \u2014 never treat one as settled." }
+          }
+        }
+      },
       {
         name: "recall_analytics_summary",
         description: `Cross-tool spend & activity overview. Returns total sessions, total cost,
@@ -27028,8 +27177,10 @@ ${JSON.stringify(r, null, 2)}
         const wal = getWAL();
         wal.log("decision_record", { subject: params.subject, importance: params.importance });
         const confidence = Math.min(1, params.importance / 5);
+        const area = canonArea(params.area);
+        const kgSubject = area ? decisionSubject(params.project, area) : params.subject;
         await remotePost("/api/kg/add", {
-          subject: params.subject,
+          subject: kgSubject,
           predicate: "decided",
           object: params.decision,
           confidence,
@@ -27037,11 +27188,12 @@ ${JSON.stringify(r, null, 2)}
         });
         if (params.reason) {
           await remotePost("/api/kg/add", {
-            subject: params.subject,
+            subject: kgSubject,
             predicate: "because",
             object: params.reason,
             confidence,
-            source_session: params.session_id
+            source_session: params.session_id,
+            supersede: false
           });
         }
         const diaryText = params.reason ? `Decided: ${params.decision}
@@ -27057,12 +27209,58 @@ Subject: ${params.subject}`;
         return {
           content: [{
             type: "text",
-            text: `Decision recorded.
-- KG: ${params.subject} \u2192 decided \u2192 ${params.decision}
-- Diary entry: ${diary.id} (importance ${params.importance})`
+            text: [
+              "Decision recorded.",
+              `- KG: ${kgSubject} \u2192 decided \u2192 ${params.decision}`,
+              area ? `- Area: ${area}${isKnownArea(area) ? "" : " (not a standard area \u2014 kept as a slug)"} \xB7 a later decision on this area supersedes this one` : "- No area given, so nothing this contradicts was closed. Pass `area` to make reversals work.",
+              `- Diary entry: ${diary.id} (importance ${params.importance})`
+            ].join("\n")
           }]
         };
       }
+      // ── The decision register ──────────────────────────────────
+      case "recall_decisions": {
+        const p = RecallDecisionsSchema.parse(args ?? {});
+        requireRemote();
+        const r = await remoteGetQS("/api/decisions", {
+          project: p.project,
+          include_candidates: p.include_candidates ? "1" : "0"
+        });
+        const rows = p.area ? r.decisions.filter((d) => d.area === canonArea(p.area)) : r.decisions;
+        const lines = [];
+        lines.push(r.project ? `Decisions for ${r.project}` : "Decisions \u2014 account-wide");
+        if (rows.length === 0) {
+          lines.push("", p.area ? `Nothing decided for "${p.area}". You are free to choose \u2014 record it afterwards with recall_decision_record.` : "Nothing decided yet.");
+        } else {
+          lines.push("");
+          for (const d of rows) {
+            const tags = [
+              d.override ? "OVERRIDES ACCOUNT" : "",
+              d.inherited ? "inherited from account" : "",
+              d.advisory ? "advisory \u2014 a personal preference, not binding" : ""
+            ].filter(Boolean).join(" \xB7 ");
+            lines.push(`${d.area}: ${d.value}${d.since ? `  (since ${d.since})` : ""}${tags ? `  [${tags}]` : ""}`);
+            if (d.why)
+              lines.push(`    why: ${d.why}`);
+            const replaced = d.history.filter((h) => !h.current);
+            for (const h of replaced)
+              lines.push(`    replaced: ${h.value} (ended ${h.to})`);
+            if (d.source_session)
+              lines.push(`    from session ${d.source_session}`);
+          }
+        }
+        if (!p.area && r.gaps.length) {
+          lines.push("", `Not decided: ${r.gaps.map((g) => g.area).join(", ")}`);
+          lines.push("Free to choose in these. Record the choice so the next session inherits it.");
+        }
+        if (p.include_candidates && r.candidates?.length) {
+          lines.push("", "Unconfirmed guesses from indexing \u2014 NOT decisions, do not treat as settled:");
+          for (const c of r.candidates.slice(0, 10)) {
+            lines.push(`  ${c.value} (seen ${c.mentions}\xD7${c.last_seen ? `, latest ${c.last_seen}` : ""})`);
+          }
+        }
+        return { content: [{ type: "text", text: lines.join("\n") }] };
+      }
       // ── Analytics summary ──────────────────────────────────────
       case "recall_analytics_summary": {
         const analyticsParams = RecallAnalyticsSummarySchema.parse(args ?? {});
@@ -27808,7 +28006,7 @@ function installCrashGuards() {
     reportClientEvent2("mcp_crash", { message: err.stack || err.message });
   });
 }
-var RELAY_VERSION = true ? "0.6.4" : "0.0.0";
+var RELAY_VERSION = true ? "0.6.5" : "0.0.0";
 function daemonIdleSecs() {
   const raw = process.env.CHAT_RECALL_DAEMON_IDLE_SECS;
   if (raw === void 0) return 900;

package/dist/watch.js

@@ -13769,12 +13769,12 @@ function extractEntities(text, context = {}) {
         if (chosen && looksLikeStrictTool(chosen) && looksLikeStrictTool(rejected)) {
           const c = canonTool(chosen), r = canonTool(rejected);
           if (projectName)
-            addTriple(projectName, "chose", c, 0.8);
-          addTriple(projectName || "project", "rejected", r, 0.7);
-          addTriple(c, "chosen_over", r, 0.8);
+            addTriple(projectName, "chose", c, DECISION_GUESS_CONFIDENCE);
+          addTriple(projectName || "project", "rejected", r, DECISION_GUESS_CONFIDENCE);
+          addTriple(c, "chosen_over", r, DECISION_GUESS_CONFIDENCE);
         }
       } else if (chosen && looksLikeDecisionObject(chosen) && projectName) {
-        addTriple(projectName, "chose", canonTool(chosen), 0.8);
+        addTriple(projectName, "chose", canonTool(chosen), DECISION_GUESS_CONFIDENCE);
       }
     }
   }
@@ -13811,7 +13811,7 @@ function extractEntities(text, context = {}) {
   }
   return triples;
 }
-var KNOWN_TOOLS, AMBIGUOUS_TOOLS, toolContextRe, TOOL_PATTERNS, DECISION_PATTERNS2, DECISION_OBJECT_STOPWORDS, KNOWN_TOOL_NAMES, CANONICAL_TOOL, canonTool, PROJECT_PATH_PATTERN, PEOPLE_PATTERNS, NOT_PEOPLE;
+var KNOWN_TOOLS, AMBIGUOUS_TOOLS, toolContextRe, TOOL_PATTERNS, DECISION_PATTERNS2, DECISION_OBJECT_STOPWORDS, KNOWN_TOOL_NAMES, CANONICAL_TOOL, canonTool, DECISION_GUESS_CONFIDENCE, PROJECT_PATH_PATTERN, PEOPLE_PATTERNS, NOT_PEOPLE;
 var init_entity_extractor = __esm({
   "../engine/dist/core/entity-extractor.js"() {
     "use strict";
@@ -14032,6 +14032,7 @@ var init_entity_extractor = __esm({
       "nextjs": "next.js"
     };
     canonTool = (n) => CANONICAL_TOOL[n.toLowerCase()] ?? n;
+    DECISION_GUESS_CONFIDENCE = 0.6;
     PROJECT_PATH_PATTERN = /\/(?:home|Users)\/\w+\/(?:code|projects|work|dev|src)\/(?:\w+\/)?([a-zA-Z][a-zA-Z0-9_-]{1,40})/g;
     PEOPLE_PATTERNS = [
       /\b(?:my name is|I'm|I am)\s+([A-Z][a-z]{1,15}(?:\s+[A-Z][a-z]{1,15})?)/g,
@@ -18304,7 +18305,7 @@ var init_http = __esm({
       CLI_VERSION = JSON.parse(readFileSync31(new URL("../package.json", import.meta.url), "utf-8")).version || CLI_VERSION;
     } catch {
       try {
-        if (true) CLI_VERSION = "0.6.4";
+        if (true) CLI_VERSION = "0.6.5";
       } catch {
       }
     }
@@ -18820,7 +18821,7 @@ var init_telemetry = __esm({
     init_sync_client();
     cliVersion = () => {
       try {
-        return true ? "0.6.4" : "dev";
+        return true ? "0.6.5" : "dev";
       } catch {
         return "dev";
       }
@@ -20763,7 +20764,7 @@ var init_sync_client = __esm({
       ownVersion = JSON.parse(readFileSync35(join42(dir, "../package.json"), "utf-8")).version || ownVersion;
     } catch {
       try {
-        if (true) ownVersion = "0.6.4";
+        if (true) ownVersion = "0.6.5";
       } catch {
       }
     }
@@ -23847,7 +23848,7 @@ try {
   ownVersion2 = JSON.parse(readFileSync39(join48(dir, "../package.json"), "utf-8")).version || ownVersion2;
 } catch {
   try {
-    if (true) ownVersion2 = "0.6.4";
+    if (true) ownVersion2 = "0.6.5";
   } catch {
   }
 }
@@ -24195,7 +24196,7 @@ async function runCollectorMigration(opts) {
 init_sync_ledger();
 init_collector_health();
 import { existsSync as existsSync49, readFileSync as readFileSync45, statSync as statSync44 } from "fs";
-var CLI_VERSION3 = true ? "0.6.4" : "0.0.0";
+var CLI_VERSION3 = true ? "0.6.5" : "0.0.0";
 var DEBOUNCE_MS = 5e3;
 var CLAUDE_DIR = claudeBackend.projectsDir();
 var PLANS_DIR = claudeBackend.plansDir();

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "chat-recall",
-  "version": "0.6.4",
+  "version": "0.6.5",
   "repository": {
     "type": "git",
     "url": "git+https://github.com/munhq/chat-recall.git"

package/skills/chat-recall/SKILL.md

@@ -29,7 +29,7 @@ All tools are MCP tools on the `chat-recall` server, named
 **Reads are safe to call proactively. Writes are not.** Everything below is a
 read except these, which change stored memory or state — treat each as an edit
 and apply the test in `chat-recall-memory` before calling one:
-`recall_kg_add`, `recall_kg_invalidate`, `recall_decision_record`,
+`recall_kg_add`, `recall_kg_invalidate`, `recall_decision_record`, `recall_decisions`,
 `recall_diary_write`, `recall_set`, `recall_task_create`, `recall_task_update`,
 `recall_task_comment`,
 `recall_task_policy`,
@@ -88,6 +88,7 @@ CLAUDE.md rule is additive and reversible, so take it; ask first before setting
 **Findings, ranked** — `recall_claude_suggestions` (every finding that becomes an agent-instruction change — the CLAUDE.md rules and skill installs — merged across account scope and every indexed project) and `recall_improvements` (everything else, most urgent first; `create_tasks: true` opens one team task per item). They split the same engines, so an item never appears in both. Neither needs the codeindex binary: without it you still get the account-level half.
 **What changed** — `recall_edits_timeline` (cross-tool edits), `recall_diff` (per-session diffs), `recall_commits` (did it actually land), `recall_markers` (where a session went sideways).
 **Durable memory / KG** — `recall_kg_query` / `recall_kg_add` / `recall_kg_invalidate` / `recall_kg_timeline` / `recall_kg_stats`, `recall_decision_record`, `recall_diary_write` / `recall_diary_read`, `recall_set` / `recall_get`.
+**Decisions** — `recall_decisions` (what is already settled per area, cascade resolved; call it before choosing a library or a datastore).
 **Team** — `recall_tasks` (the shared task board), `recall_task_create`, `recall_task_update` (status, assignee, linked session; task ids look like `t_…`; when you start work on a task, set it `in_progress` and pass your session id as `linked_session_id` so the board can verify the work shipped), `recall_task_comment` (leave a progress note on a card), `recall_task_policy` (read or change WHAT the board files — severity floor, open-card ceiling, cards per run, categories, and the projects that never file at all via `exclude_project` — reach for that one when the user says "do not put work on my board for this repo", e.g. a client's codebase they keep separate; call it with no arguments to read, and reach for it when the user says "only security", "high and critical for now", or asks why a full board has stopped filing), `recall_team_activity` (per-teammate × per-project rollup), `recall_shares` (which projects are shared — private by default, nothing is visible to teammates until shared).
 **Toolkit sync across tools and devices** — `recall_toolkit_status` (which MCP servers, skills, commands and agents exist, on which tool, on which device — and which of them can be installed elsewhere at all), `recall_toolkit_sync` (queue an install: everything or one type, every device or one machine, or fan a machine's own artifacts across its own tools). Reach for these on "set up my other laptop", "get my MCPs everywhere", "why does this tool not have my skill".
 

package/skills/chat-recall-memory/SKILL.md

@@ -26,11 +26,27 @@ future sessions inherit them.
 - `mcp__chat-recall__recall_diary_read` — prior agent diary entries (narrative
   notes across sessions). `mcp__chat-recall__recall_get` — a scoped key/value.
 
+## Check what is already decided (do this BEFORE you choose)
+
+- `mcp__chat-recall__recall_decisions` — the decision register, cascade already
+  resolved: one value per area (auth, database, payments, deploy, pricing,
+  licensing…), what it replaced and when, and the session behind it. Call it
+  before picking a library, a datastore or an auth method. If the area is
+  already decided, follow it or say plainly that you are proposing a reversal.
+  It also returns `gaps` — areas nobody has decided, where you are free to
+  choose and should record the choice afterwards.
+
+  Project overrides beat account-wide decisions. A personal preference is
+  returned marked `advisory` and never overrides a team decision.
+
 ## Record new knowledge (do this when something non-obvious is decided)
 
 - `mcp__chat-recall__recall_decision_record` — an explicit decision (subject,
-  decision, reason, importance). Shows up in future wake-up context. Use the
-  moment you and the user agree on something a later session would need.
+  decision, reason, importance). **Pass `area`**: it is the key a reversal
+  supersedes on, so recording a new `auth` decision closes the previous one and
+  leaves every other area alone. Without it nothing the decision contradicts is
+  closed. Shows up in future wake-up context. Use the moment you and the user
+  agree on something a later session would need.
 - `mcp__chat-recall__recall_kg_add` — assert a fact triple
   (subject → predicate → object, optional `valid_from`). E.g.
   `("chat-recall", "uses", "Postgres")`.