@rubytech/create-maxy-code 0.1.75 → 0.1.80

package/payload/platform/templates/specialists/agents/librarian.md

@@ -3,7 +3,7 @@ name: librarian
 description: "Foreground ingester for the memory graph. Owns external-archive and document ingestion — PDFs, web pages, plain text, conversation transcripts (WhatsApp / Telegram / Signal / iMessage / Slack / LinkedIn DMs / Zoom / meeting minutes), LinkedIn Basic Data Exports, and post-unzip trees the admin agent forwards. Loads `document-ingest`, `conversation-archive`, `conversation-archive-enrich`, and `linkedin-import` skills. Returns to the admin agent with a structured outcome line; never interacts with the operator directly except through the dispatching agent."
 summary: "Owns foreground ingest of documents, conversation transcripts, and external archives into the memory graph."
 model: claude-sonnet-4-5
-tools: mcp__memory__memory-ingest-extract, mcp__memory__memory-ingest, mcp__memory__memory-ingest-web, mcp__memory__memory-classify, mcp__memory__memory-archive-write, mcp__memory__memory-write, mcp__memory__memory-update, mcp__memory__memory-search, mcp__memory__conversation-archive-derive-insights, mcp__memory__conversation-archive-enrich-rejection, mcp__graph__maxy-graph-read_neo4j_cypher, mcp__graph__maxy-graph-write_neo4j_cypher, mcp__contacts__contact-create, mcp__tasks__task-create, Read, Bash
+tools: mcp__memory__memory-ingest-extract, mcp__memory__memory-ingest, mcp__memory__memory-ingest-web, mcp__memory__memory-classify, mcp__memory__memory-archive-write, mcp__memory__memory-write, mcp__memory__memory-update, mcp__memory__memory-search, mcp__memory__conversation-archive-derive-insights, mcp__memory__conversation-archive-enrich-rejection, mcp__graph__maxy-graph-read_neo4j_cypher, mcp__graph__maxy-graph-write_neo4j_cypher, mcp__contacts__contact-create, mcp__work__work-create, Read, Bash
 ---
 
 # Librarian

package/payload/platform/templates/specialists/agents/personal-assistant.md

@@ -32,7 +32,7 @@ Each domain has a small set of tools and, where it exists, a skill that drives t
 
 **Timezones are not optional.** Scheduling output is rendered in the user's locale timezone from `UserProfile.timezone` (IANA). Storage is UTC. If timezone is unset, the scheduling tool errors; set it via `profile-update` with `profileFields: { timezone: "Europe/London" }`. There is no UTC fallback.
 
-**Graph adjacency at write time.** Wrapped writers (`schedule-event`, `contact-create`, `task-create`) reject zero-edge calls. Resolve target elementIds via `memory-search` and pass them in the create call. The current admin session's Conversation satisfies the rule automatically when you are inside a session.
+**Graph adjacency at write time.** Wrapped writers (`schedule-event`, `contact-create`, `work-create`) reject zero-edge calls. Resolve target elementIds via `memory-search` and pass them in the create call. The current admin session's Conversation satisfies the rule automatically when you are inside a session.
 
 **WhatsApp ToS.** Automated broadcast is forbidden. The platform blocks broadcast for WhatsApp regardless of channel config; do not try to design around it.
 

package/payload/platform/templates/specialists/agents/project-manager.md

@@ -3,7 +3,7 @@ name: project-manager
 description: "Project and task management. Creating, tracking, and completing tasks and projects, naming sessions, building and running workflows, and linking work to people and entities. Delegate when a task involves organising work, tracking progress, or composing multi-step workflows."
 summary: "Manages your tasks, projects, sessions, and workflows: linking work to people and goals, and keeping everything organised."
 model: claude-sonnet-4-6
-tools: mcp__tasks__task-create, mcp__tasks__task-update, mcp__tasks__task-list, mcp__tasks__task-get, mcp__tasks__task-relate, mcp__tasks__task-complete, mcp__tasks__task-ready, mcp__tasks__project-create, mcp__tasks__project-list, mcp__tasks__project-get, mcp__tasks__project-update, mcp__tasks__project-complete, mcp__tasks__session-list, mcp__tasks__session-name, mcp__workflows__workflow-create, mcp__workflows__workflow-list, mcp__workflows__workflow-get, mcp__workflows__workflow-update, mcp__workflows__workflow-delete, mcp__workflows__workflow-validate, mcp__workflows__workflow-execute, mcp__workflows__workflow-runs, mcp__memory__memory-search
+tools: mcp__work__work-create, mcp__work__work-update, mcp__work__work-list, mcp__work__work-get, mcp__work__work-relate, mcp__work__work-complete, mcp__work__work-ready, mcp__work__project-create, mcp__work__project-list, mcp__work__project-get, mcp__work__project-update, mcp__work__project-complete, mcp__work__session-list, mcp__work__session-name, mcp__workflows__workflow-create, mcp__workflows__workflow-list, mcp__workflows__workflow-get, mcp__workflows__workflow-update, mcp__workflows__workflow-delete, mcp__workflows__workflow-validate, mcp__workflows__workflow-execute, mcp__workflows__workflow-runs, mcp__memory__memory-search
 ---
 
 # Project Manager
@@ -22,7 +22,7 @@ These three rules win when anything else in this prompt conflicts with them.
 
 Create when work is identified, do not ask. Pass real elementIds you resolved via `memory-search`: `AFFECTS` for entities the work modifies, `RAISED_BY` for the requestor, `BLOCKS` for source-blocks-target sequential prerequisites. `AFFECTS` doubles as conflict detection: two active tasks affecting the same entity surface a conflict. Priorities: urgent (this session), high (this week), normal (standard backlog), low (no deadline).
 
-Call `task-ready` before starting to check prerequisites and `AFFECTS` conflicts. Append progress to notes; never overwrite. Use `task-complete` to close, then report any tasks unblocked by the completion.
+Call `work-ready` before starting to check prerequisites and `AFFECTS` conflicts. Append progress to notes; never overwrite. Use `work-complete` to close, then report any tasks unblocked by the completion.
 
 ## Projects
 
@@ -42,7 +42,7 @@ Workflows can be triggered by scheduled events through the `check-due-events` di
 
 ## Graph adjacency at write time
 
-`task-create`, `project-create`, and `workflow-create` reject zero-edge calls. Pass at least one of: a `raisedBy` Person elementId, at least one `affects` entity elementId, or workItems (projects). When invoked inside an admin session, the session's Conversation satisfies the rule automatically. `workflow-create` requires an active admin session: the Workflow is linked `PART_OF` the session's Conversation at creation, and calls from outside a session throw.
+`work-create`, `project-create`, and `workflow-create` reject zero-edge calls. Pass at least one of: a `raisedBy` Person elementId, at least one `affects` entity elementId, or workItems (projects). When invoked inside an admin session, the session's Conversation satisfies the rule automatically. `workflow-create` requires an active admin session: the Workflow is linked `PART_OF` the session's Conversation at creation, and calls from outside a session throw.
 
 ## Domain context you do not own tools for
 

package/payload/premium-plugins/real-agent/plugins/leads/PLUGIN.md

@@ -54,7 +54,7 @@ No MCP server. Skills operate via existing platform tools:
 
 - `memory-search` to retrieve domain knowledge from the knowledge base
 - `profile-read` and `profile-update` for the customisation profile
-- `task-create` to file an open task when a chase needs deferring
+- `work-create` to file an open work item when a chase needs deferring
 
 ## References
 

package/payload/server/{chunk-ITQ55PCQ.js → chunk-TFEIFU7F.js}

@@ -256,7 +256,7 @@ var require_dist = __commonJS({
           if (matched.length === 0) {
             process.stderr.write(`[graph-write] reject reason=missing-provenance labels=${labelCsv} agent=${agentLabel}
 `);
-            throw new Error(`missing-provenance: write to ${labelCsv} requires an inbound :PRODUCED edge from a :Task, :Conversation, or :Message (createdBy.agent='${agentLabel}'). Either pass producedByTaskId on the write (autonomous workflows: call task-create at the start of the flow and thread the returned taskId), or rely on the MCP wrapper's CONVERSATION_NODE_ID env-stamp injection (direct admin asks).`);
+            throw new Error(`missing-provenance: write to ${labelCsv} requires an inbound :PRODUCED edge from a :Task, :Conversation, or :Message (createdBy.agent='${agentLabel}'). Either pass producedByTaskId on the write (autonomous workflows: call work-create at the start of the flow and thread the returned taskId), or rely on the MCP wrapper's CONVERSATION_NODE_ID env-stamp injection (direct admin asks).`);
           }
           provenanceSourceId = matched[0].rel.targetNodeId;
           provenanceSourceLabel = matched[0].sourceLabel;

package/payload/server/{chunk-HZ77KMTV.js → chunk-W33UUYYR.js}

@@ -3633,6 +3633,7 @@ function walkPremiumBundles() {
   }
   const result = [];
   for (const bundle of entries) {
+    if (bundle.startsWith(".")) continue;
     const bundleDir = resolve4(stagingRoot, bundle);
     try {
       if (!statSync2(bundleDir).isDirectory()) continue;
@@ -3738,9 +3739,6 @@ function reconcileEnabledPlugins(accountDir, config) {
 // app/lib/claude-agent/logging.ts
 var emitMissingOnResolve = (..._args) => null;
 
-// app/lib/claude-agent/spawn-env.ts
-var getBundleMtimeIso = (..._args) => null;
-
 // app/lib/claude-agent/summary-helpers.ts
 var stripAttachmentMetaSuffix = (..._args) => null;
 
@@ -4435,7 +4433,6 @@ export {
   autoDeliverPremiumPlugins,
   reconcileEnabledPlugins,
   emitMissingOnResolve,
-  getBundleMtimeIso,
   stripAttachmentMetaSuffix,
   requirePortEnv,
   safeJson,

package/payload/server/{cloudflare-task-tracker-PZBX7X3D.js → cloudflare-task-tracker-7P355PSD.js}

@@ -7,7 +7,7 @@ import {
   readTunnelState,
   recoverRunningCloudflareTasks,
   resolveUnitGoneVerdict
-} from "./chunk-ITQ55PCQ.js";
+} from "./chunk-TFEIFU7F.js";
 import "./chunk-VGMVEYX7.js";
 import "./chunk-JSBRDJBE.js";
 export {

package/payload/server/maxy-edge.js

@@ -19,7 +19,7 @@ import {
   streamActionEvents,
   vncLog,
   websockifyLog
-} from "./chunk-HZ77KMTV.js";
+} from "./chunk-W33UUYYR.js";
 import "./chunk-VGMVEYX7.js";
 import "./chunk-JSBRDJBE.js";
 

package/payload/server/server.js

@@ -25,7 +25,6 @@ import {
   findMissingPlugins,
   fingerprintSessionKey,
   getAccountIdForSession,
-  getBundleMtimeIso,
   getConversationIdForSession,
   getRoleForSession,
   getSessionKeyByConversationId,
@@ -60,7 +59,7 @@ import {
   vncLog,
   waitForExit,
   walkPremiumBundles
-} from "./chunk-HZ77KMTV.js";
+} from "./chunk-W33UUYYR.js";
 import {
   CLOUDFLARE_TASK_DIAGNOSTICS,
   appendCloudflareSteps,
@@ -69,7 +68,7 @@ import {
   openCloudflareTask,
   readTunnelState,
   resolveUnitGoneVerdict
-} from "./chunk-ITQ55PCQ.js";
+} from "./chunk-TFEIFU7F.js";
 import {
   GREETING_DIRECTIVE,
   deleteAgentProjection,
@@ -8059,6 +8058,12 @@ async function performSpawnWithInitialMessage(args) {
   console.log(`${TAG18} tunnel-url-resolved value=${tunnelUrl ?? "null"}`);
   const upstreamPayload = JSON.stringify({
     senderId: args.senderId,
+    // Task 205 — pass userId through to the manager so it lands as
+    // `USER_ID` in the pty child env and as `${USER_ID}` in the
+    // per-spawn mcp.json placeholder map. The cookie-auth path
+    // resolves it from `getUserIdForSession`; the Stop-hook recorder
+    // path legitimately leaves it undefined.
+    userId: args.userId,
     role: args.role,
     channel: args.channel,
     permissionMode: args.permissionMode,
@@ -8215,6 +8220,9 @@ app14.post("/resume", async (c) => {
     headers: { "content-type": "application/json" },
     body: JSON.stringify({
       senderId,
+      // Task 205 — resume parity with /spawn: thread userId so the
+      // respawned pty inherits the same USER_ID env.
+      userId,
       role: "admin",
       channel,
       sessionId: body.sessionId,
@@ -8900,6 +8908,34 @@ var CLOUDFLARE_SETUP_FORM_SCHEMA = {
   secretFields: ["password", "session_key", "messageId"]
 };
 
+// app/lib/cloudflare-bundle-mtime.ts
+import { statSync as statSync8 } from "fs";
+var cachedBundleMtimeIso;
+function getBundleMtimeIso(entry = process.argv[1]) {
+  if (cachedBundleMtimeIso !== void 0) return cachedBundleMtimeIso;
+  if (!entry) {
+    process.stderr.write(`[bundle-mtime] fallback reason=empty-path
+`);
+    cachedBundleMtimeIso = "unknown";
+    return cachedBundleMtimeIso;
+  }
+  try {
+    cachedBundleMtimeIso = statSync8(entry).mtime.toISOString();
+  } catch (e) {
+    const reason = e instanceof Error ? e.message : String(e);
+    process.stderr.write(`[bundle-mtime] fallback reason=stat-threw error="${reason}"
+`);
+    cachedBundleMtimeIso = "unknown";
+  }
+  return cachedBundleMtimeIso;
+}
+function decideRetryBranch(args) {
+  const { bundleMtimeIso, prior } = args;
+  if (!prior || prior.outcome !== "failed") return "proceed";
+  if (bundleMtimeIso === "unknown") return "surface-error";
+  return bundleMtimeIso > prior.completedAt ? "re-invoke" : "surface-error";
+}
+
 // server/routes/admin/cloudflare.ts
 function cloudflareStreamLogPath(accountId, correlationId) {
   return resolve14(ACCOUNTS_DIR, accountId, "cloudflare-stream-logs", `${correlationId}.log`);
@@ -9282,16 +9318,12 @@ app20.post("/setup", requireAdminSession, async (c) => {
   const bundleMtimeIso = getBundleMtimeIso();
   try {
     const prior = await findMostRecentTerminalCloudflareTask(accountId, correlationId);
-    if (prior && prior.outcome === "failed") {
-      const bundleNewer = bundleMtimeIso !== "unknown" && bundleMtimeIso > prior.completedAt;
-      const branch = bundleNewer ? "re-invoke" : "surface-error";
-      log(`retry-decision bundleMtime=${bundleMtimeIso} lastErrorAt=${prior.completedAt} branch=${branch}`);
-      if (branch === "surface-error") {
-        const literal = prior.errorMessage ?? CLOUDFLARE_TASK_DIAGNOSTICS.scriptExitedNonzero;
-        return err("script", literal);
-      }
-    } else {
-      log(`retry-decision bundleMtime=${bundleMtimeIso} lastErrorAt=none branch=proceed`);
+    const branch = decideRetryBranch({ bundleMtimeIso, prior });
+    const lastErrorAt = prior && prior.outcome === "failed" ? prior.completedAt : "none";
+    log(`retry-decision bundleMtime=${bundleMtimeIso} lastErrorAt=${lastErrorAt} branch=${branch}`);
+    if (branch === "surface-error" && prior) {
+      const literal = prior.errorMessage ?? CLOUDFLARE_TASK_DIAGNOSTICS.scriptExitedNonzero;
+      return err("script", literal);
     }
   } catch (e) {
     log(`retry-decision lookup-failed reason="${e instanceof Error ? e.message : String(e)}" branch=proceed`);
@@ -12252,7 +12284,7 @@ app33.route("/health-brand", health_default2);
 var admin_default = app33;
 
 // server/routes/sites.ts
-import { existsSync as existsSync20, readFileSync as readFileSync18, realpathSync as realpathSync5, statSync as statSync8 } from "fs";
+import { existsSync as existsSync20, readFileSync as readFileSync18, realpathSync as realpathSync5, statSync as statSync9 } from "fs";
 import { resolve as resolve21 } from "path";
 var SAFE_SEG_RE = /^[a-z0-9_][a-z0-9_.-]{0,99}$/i;
 var MIME = {
@@ -12318,7 +12350,7 @@ app34.get("/:rel{.*}", (c) => {
   }
   let stat7;
   try {
-    stat7 = existsSync20(filePath) ? statSync8(filePath) : null;
+    stat7 = existsSync20(filePath) ? statSync9(filePath) : null;
   } catch {
     stat7 = null;
   }
@@ -12502,7 +12534,7 @@ function startReaper2() {
 
 // ../lib/entitlement/src/index.ts
 import { createPublicKey, createHash as createHash3, verify as cryptoVerify } from "crypto";
-import { existsSync as existsSync21, readFileSync as readFileSync19, statSync as statSync9 } from "fs";
+import { existsSync as existsSync21, readFileSync as readFileSync19, statSync as statSync10 } from "fs";
 import { resolve as resolve22 } from "path";
 
 // ../lib/entitlement/src/canonicalize.ts
@@ -12554,7 +12586,7 @@ function resolveEntitlement(brand, account) {
   if (!existsSync21(entitlementPath)) {
     return logResolved(anonymousFallback("missing"), { reason: "missing" });
   }
-  const stat7 = statSync9(entitlementPath);
+  const stat7 = statSync10(entitlementPath);
   const key = memoKey(stat7.mtimeMs, account);
   if (memo && memo.key === key) {
     return memo.result;
@@ -13522,7 +13554,7 @@ reconcileEnabledPlugins(bootAccount?.accountDir, bootAccount?.config);
 (async () => {
   if (!bootAccount) return;
   try {
-    const { recoverRunningCloudflareTasks } = await import("./cloudflare-task-tracker-PZBX7X3D.js");
+    const { recoverRunningCloudflareTasks } = await import("./cloudflare-task-tracker-7P355PSD.js");
     const result = await recoverRunningCloudflareTasks(
       bootAccount.accountId,
       configDirForWhatsApp,

package/payload/platform/plugins/tasks/PLUGIN.md

@@ -1,120 +0,0 @@
----
-name: tasks
-description: "Graph-backed task lifecycle management. Task nodes in Neo4j with dependency tracking (BLOCKS), conflict detection (AFFECTS), and a parallel safety query (task-ready). Surfaced by memory-search."
-tools:
-  - name: task-create
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-update
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-list
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-get
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-relate
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-complete
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: task-ready
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: session-list
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: session-name
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: project-create
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: project-list
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: project-get
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: project-update
-    publicAllowlist: false
-    adminAllowlist: false
-  - name: project-complete
-    publicAllowlist: false
-    adminAllowlist: false
-metadata: {"platform":{"always":false,"embed":[],"pluginKey":"tasks"}}
-mcp:
-  command: node
-  args:
-    - ${PLATFORM_ROOT}/lib/mcp-spawn-tee/dist/index.js
-    - ${PLATFORM_ROOT}/plugins/tasks/mcp/dist/index.js
-  env:
-    MCP_SPAWN_TEE_NAME: tasks
-    LOG_DIR: ${LOG_DIR}
-    PLATFORM_ROOT: ${PLATFORM_ROOT}
-    ACCOUNT_ID: ${ACCOUNT_ID}
-    SESSION_ID: ${SESSION_ID}
----
-
-# Tasks
-
-Tasks live in the graph — not in files, not in memory. When the user identifies a piece of work, create a Task node immediately and relate it to the entities it will modify. Surface open tasks at the start of relevant conversations.
-
-## Creating Tasks
-
-Create a task when the user states something that needs doing, when a conversation reveals a follow-up, or when a business obligation or deadline is mentioned. Don't ask — create it and confirm.
-
-When creating, link the task to the entities it will modify via `AFFECTS` relationships. If the user mentions who requested it, link via `RAISED_BY`. If the task has prerequisites ("after the materials arrive"), create a `BLOCKS` edge from the prerequisite to this task.
-
-## Dependencies and Conflicts
-
-**BLOCKS** — sequential dependency. The target task cannot start until the source is completed. When the user says "start the kitchen refit," check for blockers and explain what's outstanding.
-
-**AFFECTS** — conflict detection. Two active tasks that both `AFFECTS` the same entity are in conflict. Surface this when either task is started: "This invoice is also being modified by [other task] — resolve that first or confirm you want both in progress."
-
-## Before Starting a Task
-
-Check `task-ready` before moving a task to `active`. If the task has unresolved blockers, explain what's outstanding. If another active task `AFFECTS` the same entity, surface the conflict and let the user resolve it.
-
-## During a Task
-
-Append progress to notes — never overwrite existing history. If the task creates or modifies graph entities, the `AFFECTS` relationships should already exist from creation. If not, add them.
-
-## Completing a Task
-
-Use `task-complete` to mark done. It re-embeds the task and surfaces any tasks that were blocked by this one and are now ready. Proactively tell the user what's been unlocked: "The materials order is complete — the kitchen refit is now unblocked."
-
-## Session Start
-
-Run `task-ready` and surface the ready set if the user has open tasks. This answers "what can I work on next without conflict?"
-
-## Priority
-
-- `urgent` — must be handled in this session
-- `high` — should happen this week
-- `normal` — standard backlog
-- `low` — nice to have, no deadline pressure
-
-## Tools
-
-- `task-create` — Create a task, optionally linked to a person (RAISED_BY) and entities (AFFECTS). For durable-action records, pass `kind` (e.g. `"onboarding-establish-owner"`, `"cloudflare-tunnel-login"`) to set status `running`; pass `inputs` with the form payload and `inputSchema` declaring any secret fields — values land on the Task as `inputs.<field>` props after `redactSecrets` strips schema-tagged keys; `inputsProvided` records the call-shape (names submitted) alongside the value-bearing `inputs.*`. Pass `raisedDuringConversationKey` to write a `:RAISED_DURING` edge to the originating Conversation when it differs from the current session.
-- `task-update` — Update name, description, status, priority, due date, or append a note (append-only). Pass `appendStep` to atomically append one entry to `steps[]` (call once per action phase); `errorMessage` is required when status transitions to `failed`; `completed` and `failed` transitions auto-set `completedAt`.
-- `task-list` — List tasks filtered by status, priority, or related entity
-- `task-get` — Fetch a task with full details and all relationships
-- `task-relate` — Add AFFECTS, BLOCKS, PART_OF, or RAISED_BY relationships. For Task targets, pass the taskId (UUID) from task-create. For other labels, pass the elementId from memory-search.
-- `task-complete` — Mark done, re-embed, surface newly-unblocked tasks
-- `task-ready` — Safe-to-start set: pending tasks with no unresolved BLOCKS and no AFFECTS conflicts
-- `session-list` — List recent sessions with their linked tasks
-- `session-name` — Set a human-readable name on the current session
-
-## Project Tools
-
-Project tools manage `:Project:Task` nodes — multi-label nodes with structured health tracking, lifecycle phases, and parent-child relationships via `HAS_TASK`. Use these instead of manually choreographing `task-create` + `task-relate` sequences for project work.
-
-- `project-create` — Atomic project creation: parent node, child work items, dependencies, and relationships in a single transaction. Accepts tier (quick/standard/full), optional clientRef, targetDate, workItems with blocks dependencies. Max 50 work items.
-- `project-list` — Active projects with structured health data (green/amber/red/grey signal, completion counts, overdue/blocked counts, next action). Defaults to pending + active projects.
-- `project-get` — Full project detail: children with blocker status, health, relationships, lifecycle notes, timestamps.
-- `project-update` — Update project phase, tier, targetDate, clientRef, status, or append a note. Phase changes auto-append `[PROJECT:PHASE]` lifecycle notes.
-- `project-complete` — Complete a project. Returns incomplete children so you can confirm with the user before finalising. Always completes — does not gate on unfinished work.