nexus-agents 2.137.0 → 2.138.1

package/README.md

@@ -99,20 +99,22 @@ Add to your Claude Desktop configuration (`~/Library/Application Support/Claude/
 
 The Orchestrator agent analyzes incoming tasks and delegates to specialized experts:
 
-| Expert                    | Specialization                                         |
-| ------------------------- | ------------------------------------------------------ |
-| **Code Expert**           | Implementation, debugging, optimization, refactoring   |
-| **Architecture Expert**   | System design, patterns, trade-offs, scalability       |
-| **Security Expert**       | Vulnerability analysis, secure coding, threat modeling |
-| **Documentation Expert**  | Technical writing, API docs, code comments             |
-| **Testing Expert**        | Test strategies, coverage analysis, test generation    |
-| **DevOps Expert**         | CI/CD, deployment, containerization                    |
-| **Research Expert**       | Literature review, state-of-the-art analysis           |
-| **PM Expert**             | Product management, requirements, priorities           |
-| **UX Expert**             | User experience, usability, accessibility              |
-| **Infrastructure Expert** | Server management, bare metal, networking              |
-
-Experts can collaborate on complex tasks. The Orchestrator combines their outputs into a single response.
+| Expert                        | Specialization                                         |
+| ----------------------------- | ------------------------------------------------------ |
+| **Code Expert**               | Implementation, debugging, optimization, refactoring   |
+| **Architecture Expert**       | System design, patterns, trade-offs, scalability       |
+| **Security Expert**           | Vulnerability analysis, secure coding, threat modeling |
+| **Documentation Expert**      | Technical writing, API docs, code comments             |
+| **Testing Expert**            | Test strategies, coverage analysis, test generation    |
+| **DevOps Expert**             | CI/CD, deployment, containerization                    |
+| **Research Expert**           | Literature review, state-of-the-art analysis           |
+| **PM Expert**                 | Product management, requirements, priorities           |
+| **UX Expert**                 | User experience, usability, accessibility              |
+| **Infrastructure Expert**     | Server management, bare metal, networking              |
+| **QA Expert**                 | Quality gates, test-plan review, defect triage         |
+| **Data-Visualization Expert** | Charts, diagrams, dashboards, data presentation        |
+
+All twelve expert types can collaborate on complex tasks. The Orchestrator combines their outputs into a single response.
 
 ### Model Adapters
 
@@ -155,7 +157,7 @@ steps:
 
 ### MCP Tools
 
-The server exposes 42 MCP tools for integration. Key tools include:
+The server exposes 46 MCP tools for integration. Key tools include:
 
 | Tool                 | Description                                    |
 | -------------------- | ---------------------------------------------- |
@@ -171,7 +173,7 @@ The server exposes 42 MCP tools for integration. Key tools include:
 | `repo_analyze`       | Analyze GitHub repository structure            |
 | `repo_security_plan` | Generate security scanning pipeline for a repo |
 
-See the root [README](../../README.md) for the complete tool list.
+See the [root README](https://github.com/nexus-substrate/nexus-agents/blob/main/README.md) for the complete tool list.
 
 ---
 
@@ -204,7 +206,7 @@ nexus-agents/
 └── pnpm-workspace.yaml
 ```
 
-See [docs/architecture/README.md](../../docs/architecture/README.md) for detailed module descriptions.
+See [docs/architecture/README.md](https://github.com/nexus-substrate/nexus-agents/blob/main/docs/architecture/README.md) for detailed module descriptions.
 
 ### Dependency Flow
 
@@ -280,7 +282,7 @@ import {
 // Create individual adapters
 const claude = createClaudeAdapter({ model: 'claude-sonnet-4-6' });
 const openai = createOpenAIAdapter({ model: 'gpt-4o' });
-const gemini = createGeminiAdapter({ model: 'gemini-1.5-pro' });
+const gemini = createGeminiAdapter({ model: 'gemini-3-pro' });
 const ollama = createOllamaAdapter({ model: 'llama3:8b' });
 
 // Or use the factory
@@ -379,11 +381,11 @@ We welcome contributions! Please see our guidelines:
 
 - Files must be under 400 lines
 - Functions must be under 50 lines
-- Test coverage must be at least 80%
+- Coverage thresholds: 60% statements/functions/lines, 50% branches (`vitest.config.ts`)
 - All code must pass linting and type checking
 
-See [CODING_STANDARDS.md](../../CODING_STANDARDS.md) for detailed guidelines.
-See [CONTRIBUTING.md](../../CONTRIBUTING.md) for contribution workflow.
+See [CODING_STANDARDS.md](https://github.com/nexus-substrate/nexus-agents/blob/main/CODING_STANDARDS.md) for detailed guidelines.
+See [CONTRIBUTING.md](https://github.com/nexus-substrate/nexus-agents/blob/main/CONTRIBUTING.md) for contribution workflow.
 
 ### Commit Convention
 
@@ -402,7 +404,7 @@ chore(scope): maintenance tasks
 
 ## License
 
-MIT - See [LICENSE](../../LICENSE) for details.
+MIT - See [LICENSE](https://github.com/nexus-substrate/nexus-agents/blob/main/LICENSE) for details.
 
 ---
 

package/dist/{chunk-DPHEFR5X.js → chunk-BQA5LM72.js}

@@ -42,7 +42,7 @@ import {
 } from "./chunk-DHVMSIT5.js";
 
 // src/version.ts
-var VERSION = true ? "2.137.0" : "dev";
+var VERSION = true ? "2.138.1" : "dev";
 
 // src/config/schemas-core.ts
 import { z } from "zod";
@@ -2132,7 +2132,7 @@ async function runDoctorFix(result) {
   writeLine2("\u2500".repeat(40));
   let fixCount = 0;
   if (!result.dataDirectory.rootExists || result.dataDirectory.subdirectories.some((d) => !d.exists || !d.writable)) {
-    const { runSetup } = await import("./setup-command-TIGCMJRZ.js");
+    const { runSetup } = await import("./setup-command-I3WAJY2X.js");
     const setupResult = runSetup({
       skipMcp: true,
       skipRules: true,
@@ -2245,4 +2245,4 @@ export {
   startStdioServer,
   closeServer
 };
-//# sourceMappingURL=chunk-DPHEFR5X.js.map
+//# sourceMappingURL=chunk-BQA5LM72.js.map

package/dist/{chunk-23J7J37D.js → chunk-D5FEGFJM.js}

@@ -18,7 +18,7 @@ import {
   DEFAULT_TASK_TTL_MS,
   DEFAULT_TOOL_RATE_LIMITS,
   clampTaskTtl
-} from "./chunk-DPHEFR5X.js";
+} from "./chunk-BQA5LM72.js";
 import {
   executeExpert
 } from "./chunk-7QR2RTA4.js";
@@ -40,7 +40,7 @@ import {
   toJobSummary,
   warnIfSimulatedOutsideTests,
   writeJobCancelled
-} from "./chunk-HC5CUX6C.js";
+} from "./chunk-NWW7HLRL.js";
 import {
   normalizeTopicToCanonical,
   synthesizeResearch
@@ -40046,32 +40046,6 @@ var TierTransitionPayloadSchema = z85.object({
   ratificationVoteRef: z85.string().min(1).optional()
 }).strict();
 var TIER_TRANSITION_METADATA_KEY = "tierTransition";
-var RatificationVoteSchema = z85.object({
-  /** The ref a tier-transition's `ratificationVoteRef` must equal to resolve. */
-  id: z85.string().min(1),
-  /** The loop/strategy the vote ratified (cross-checked against the transition subject). */
-  subject: z85.string().min(1),
-  /**
-   * The recorded decision. Only `approved` ratifies a promotion; `rejected`
-   * fails the gate (`promotion-ratification-not-approved`).
-   */
-  decision: z85.enum(["approved", "rejected"]),
-  /**
-   * The voting strategy. A promotion is governance-of-the-governor and must be
-   * ratified by a `higher_order` consensus_vote (ADR-0017).
-   */
-  strategy: z85.enum(["higher_order", "simple_majority", "supermajority", "unanimous"]),
-  /** ISO-8601 timestamp the vote was recorded. */
-  votedAt: z85.string().min(1),
-  /** Optional approval fraction / quorum detail, for the human record. */
-  approvalPercentage: z85.number().min(0).max(100).optional(),
-  /** Optional link to the vote artifact / issue thread for the human record. */
-  voteUri: z85.string().min(1).optional()
-}).strict();
-var RatificationVoteLedgerSchema = z85.object({
-  version: z85.number().int().positive(),
-  votes: z85.array(RatificationVoteSchema)
-}).strict();
 var AuditSeveritySchema = z85.enum([
   "info",
   // Normal operations
@@ -43409,7 +43383,7 @@ ${contextBlock}`;
       const strategy = config.votingStrategy ?? "higher_order";
       await postProgress(config, "Vote", `Running consensus with ${strategy} strategy...`);
       try {
-        const { executeVoting } = await import("./consensus-vote-HHA73ME6.js");
+        const { executeVoting } = await import("./consensus-vote-PCIFXILX.js");
         const votingResult = await executeVoting(
           {
             proposal: buildVoteProposal(plan, research),
@@ -46633,6 +46607,13 @@ function buildDefaultExecutors(trustTier) {
   return {
     "dev-pipeline": (_decision, metaInput) => runDevPipelineForGoal(metaInput.goal, trustTier),
     pipeline: (_decision, metaInput) => runPipelineForGoal(metaInput.goal),
+    // #3988: `research` deliberately ALIASES the `pipeline` engine — it runs the
+    // SAME generic stage registry (selectStageRegistry only branches greenfield/
+    // audit), shaped by the goal text, NOT a distinct research stage registry.
+    // The registry's research `entrypointTool` already points at run_pipeline, so
+    // this is intended aliasing, not a distinct executor. A real research-shaped
+    // registry is a feature with no current consumer (YAGNI) — add it only when a
+    // named loop needs research-specific stages; until then research==pipeline.
     research: (_decision, metaInput) => runPipelineForGoal(metaInput.goal),
     consensus: (_decision, metaInput) => runConsensusForGoal(metaInput.goal)
   };
@@ -50911,4 +50892,4 @@ export {
   shutdownFeedbackSubscriber,
   createEventBusBridge
 };
-//# sourceMappingURL=chunk-23J7J37D.js.map
+//# sourceMappingURL=chunk-D5FEGFJM.js.map