@harness-engineering/cli 2.8.0 → 3.0.0

package/dist/agents/skills/gemini-cli/outcome-eval/skill.yaml

@@ -0,0 +1,51 @@
+name: outcome-eval
+version: '0.1.0'
+description: >-
+  LLM-judgment skill that produces a structured, confidence-rated verdict on
+  whether an implementation satisfied its spec. Reads the spec's acceptance
+  section, the change diff, and test output; emits an OutcomeVerdict
+  (SATISFIED | NOT_SATISFIED | INCONCLUSIVE) with confidence, rationale, and
+  unmet criteria. Authority is derived in TypeScript, never from the LLM: a
+  high-confidence NOT_SATISFIED blocks ship; every other verdict is advisory.
+  The verdict persists as an execution_outcome node and feeds skill-effectiveness
+  baselines. The harness's first blocking post-execution spec-satisfaction gate.
+stability: draft
+cognitive_mode: constructive-architect
+triggers:
+  - manual
+  - on_pr
+platforms:
+  - claude-code
+tools:
+  - Bash
+  - Read
+  - Glob
+  - Grep
+mcp:
+  tool: outcome_eval
+  # specPath, diff, and testOutput are ALL required. diff/testOutput are the
+  # evidence the judge reasons over: omitting them degrades the verdict to
+  # INCONCLUSIVE/advisory (never blocking), so the agent MUST supply them.
+  input:
+    specPath: 'string (required) — absolute or repo-relative path to the spec markdown to judge against'
+    diff: 'string (required) — unified diff of the change under judgment (git diff); omission degrades the verdict to INCONCLUSIVE/advisory'
+    testOutput: 'string (required) — captured test-runner stdout+stderr; omission degrades the verdict to INCONCLUSIVE/advisory'
+    model: 'string (optional) — model override for the outcome-eval LLM call'
+    path: 'string (optional) — project root used to resolve the knowledge graph (default cwd)'
+type: rigid
+tier: 2
+phases:
+  - name: gather
+    description: Collect the unified diff (git diff), capture test output, resolve the spec path
+    required: true
+  - name: resolve
+    description: Resolve the judgment section (Success Criteria -> User-Visible Behavior -> Overview)
+    required: true
+  - name: judge
+    description: Invoke OutcomeEvaluator.evaluate(); LLM returns verdict/confidence/rationale/unmetCriteria; authority derived in TS
+    required: true
+  - name: gate
+    description: Render the verdict; on a blocking verdict (NOT_SATISFIED + high confidence) halt before Ship
+    required: true
+state:
+  persistent: false

package/dist/agents/skills/tests/harness-test-advisor.test.ts

@@ -0,0 +1,94 @@
+// agents/skills/tests/harness-test-advisor.test.ts
+//
+// Contract tests for the harness-test-advisor skill. Locks in the Coverage
+// Audit mode contract from GH issue 488: skill exposes an audit entry point,
+// SKILL.md documents the three audit phases (INVENTORY, QUALITY REVIEW,
+// GAP REPORT), and the canary plugin agents are named as the remediation
+// path. Generic schema/structure/parity checks live in the sibling
+// *.test.ts files.
+
+import { describe, it, expect } from 'vitest';
+import { readFileSync } from 'fs';
+import { resolve, dirname } from 'path';
+import { fileURLToPath } from 'url';
+import { parse } from 'yaml';
+
+const __dirname = dirname(fileURLToPath(import.meta.url));
+const SKILLS_DIR = resolve(__dirname, '..');
+
+const SKILL_NAME = 'harness-test-advisor';
+const PLATFORMS = ['claude-code', 'gemini-cli', 'cursor', 'codex'] as const;
+
+function readSkillYaml(platform: string): Record<string, unknown> {
+  const path = resolve(SKILLS_DIR, platform, SKILL_NAME, 'skill.yaml');
+  return parse(readFileSync(path, 'utf-8'));
+}
+
+function readSkillMd(platform: string): string {
+  return readFileSync(resolve(SKILLS_DIR, platform, SKILL_NAME, 'SKILL.md'), 'utf-8');
+}
+
+describe('harness-test-advisor Coverage Audit metadata', () => {
+  it.each(PLATFORMS)('%s skill.yaml exposes an audit CLI arg', (platform) => {
+    const meta = readSkillYaml(platform) as {
+      cli: { args: Array<{ name: string }> };
+    };
+    const argNames = meta.cli.args.map((a) => a.name);
+    expect(argNames).toContain('audit');
+  });
+
+  it.each(PLATFORMS)(
+    '%s skill.yaml description advertises coverage audit alongside selection',
+    (platform) => {
+      const meta = readSkillYaml(platform) as { description: string };
+      expect(meta.description).toMatch(/coverage/i);
+    }
+  );
+});
+
+describe('harness-test-advisor SKILL.md surfaces Coverage Audit mode', () => {
+  it.each(PLATFORMS)('%s SKILL.md advertises Coverage Audit in When to Use', (platform) => {
+    const body = readSkillMd(platform);
+    const whenToUse = body.split('## Prerequisites')[0] ?? '';
+    expect(whenToUse).toMatch(/Coverage Audit/);
+  });
+
+  const AUDIT_PHASES = ['INVENTORY', 'QUALITY REVIEW', 'GAP REPORT'];
+
+  it.each(PLATFORMS)('%s SKILL.md declares the three Coverage Audit phases', (platform) => {
+    const body = readSkillMd(platform);
+    for (const phase of AUDIT_PHASES) {
+      expect(body, `missing Coverage Audit phase "${phase}" in ${platform}`).toContain(phase);
+    }
+  });
+
+  const CANARY_SKILLS = [
+    'canary:canary-review-test',
+    'canary:canary-write-test',
+    'canary:canary-pick-framework',
+  ];
+
+  it.each(PLATFORMS)('%s SKILL.md routes audit follow-ups to canary plugin skills', (platform) => {
+    const body = readSkillMd(platform);
+    for (const skill of CANARY_SKILLS) {
+      expect(body, `missing canary skill reference "${skill}" in ${platform}`).toContain(skill);
+    }
+  });
+});
+
+describe('harness-test-advisor wires the canary MCP tools (deterministic path)', () => {
+  // The audit probes the canary CLI for graceful degradation and uses the
+  // deterministic framework recommender — separate from the generative plugin path.
+  const CANARY_MCP_TOOLS = ['canary_probe', 'canary_recommend_framework'];
+
+  it.each(PLATFORMS)('%s SKILL.md declares an audit PROBE phase', (platform) => {
+    expect(readSkillMd(platform)).toMatch(/Audit Phase 0: PROBE/);
+  });
+
+  it.each(PLATFORMS)('%s SKILL.md references both canary MCP tools', (platform) => {
+    const body = readSkillMd(platform);
+    for (const tool of CANARY_MCP_TOOLS) {
+      expect(body, `missing canary MCP tool "${tool}" in ${platform}`).toContain(tool);
+    }
+  });
+});

package/dist/{agents-md-WMJO5HSM.js → agents-md-XOJE5ETA.js}

@@ -1,7 +1,7 @@
 import {
   generateAgentsMd
-} from "./chunk-7EQAZKJ5.js";
-import "./chunk-GOHWCHCS.js";
+} from "./chunk-QBC7DI4Q.js";
+import "./chunk-N24HCQA4.js";
 import "./chunk-N55WOGN3.js";
 import "./chunk-BDGTZRZ6.js";
 import "./chunk-PP6ZRL5T.js";

package/dist/{architecture-EVYVGPRZ.js → architecture-JEZPJ725.js}

@@ -1,11 +1,11 @@
 import {
   checkDependenciesDefinition,
   handleCheckDependencies
-} from "./chunk-UJFNIWC7.js";
-import "./chunk-HRSE6IQD.js";
+} from "./chunk-527G27VI.js";
+import "./chunk-DE5U6KOL.js";
 import "./chunk-RC3OI5NK.js";
 import "./chunk-W6Y7ZW3Y.js";
-import "./chunk-GOHWCHCS.js";
+import "./chunk-N24HCQA4.js";
 import "./chunk-N55WOGN3.js";
 import "./chunk-BDGTZRZ6.js";
 import "./chunk-PP6ZRL5T.js";

package/dist/{assess-project-KZG563SO.js → assess-project-RHXOQNMM.js}

@@ -2,7 +2,7 @@ import {
   assessProjectDefinition,
   handleAssessProject,
   parseToolResponse
-} from "./chunk-FZ5MBXEG.js";
+} from "./chunk-YQSYCBAH.js";
 import "./chunk-RC3OI5NK.js";
 import "./chunk-W6Y7ZW3Y.js";
 import "./chunk-KFQGP6VL.js";

package/dist/bin/harness-mcp.js

@@ -1,30 +1,30 @@
 #!/usr/bin/env node
 import {
   startServer
-} from "../chunk-CAQPSFPT.js";
-import "../chunk-KBCPELBC.js";
-import "../chunk-LADPYELQ.js";
-import "../chunk-FZ5MBXEG.js";
-import "../chunk-ACOOG3UW.js";
-import "../chunk-UUO4K7NX.js";
+} from "../chunk-WWEVS7RO.js";
+import "../chunk-HEIMJJI4.js";
+import "../chunk-XX4OLNWQ.js";
+import "../chunk-YQSYCBAH.js";
+import "../chunk-Z4AWEIBY.js";
+import "../chunk-ONCPJGHY.js";
 import "../chunk-IOW3MW2K.js";
-import "../chunk-UXPWSV3B.js";
+import "../chunk-BYVT5LVO.js";
 import "../chunk-3ISINLYT.js";
 import "../chunk-CLI4K2CH.js";
-import "../chunk-KI6PHH6Q.js";
-import "../chunk-UJFNIWC7.js";
-import "../chunk-HRSE6IQD.js";
-import "../chunk-C7J55C6E.js";
+import "../chunk-HWTUUQOW.js";
+import "../chunk-527G27VI.js";
+import "../chunk-DE5U6KOL.js";
+import "../chunk-MGUPSE6D.js";
 import "../chunk-RC3OI5NK.js";
 import "../chunk-W6Y7ZW3Y.js";
 import "../chunk-27AJKSQY.js";
 import "../chunk-6B6UN6SG.js";
 import "../chunk-HIK6OEUF.js";
-import "../chunk-B7KMTHN3.js";
+import "../chunk-GGKRA7A7.js";
 import "../chunk-ZCZD7FHU.js";
 import "../chunk-EBJQ6N4M.js";
 import "../chunk-3WGJMBKH.js";
-import "../chunk-GOHWCHCS.js";
+import "../chunk-N24HCQA4.js";
 import "../chunk-N55WOGN3.js";
 import "../chunk-BDGTZRZ6.js";
 import "../chunk-PP6ZRL5T.js";

package/dist/bin/harness.js

@@ -3,38 +3,38 @@ import {
   createProgram,
   installCommandTelemetry,
   printFirstRunWelcome
-} from "../chunk-4BDOSCNY.js";
-import "../chunk-5XL5QJ5B.js";
-import "../chunk-7EQAZKJ5.js";
-import "../chunk-STESM73J.js";
-import "../chunk-VBNJAKUH.js";
-import "../chunk-EEFRPEWT.js";
+} from "../chunk-XFU6SB2Q.js";
+import "../chunk-6WQCLCBO.js";
+import "../chunk-QBC7DI4Q.js";
+import "../chunk-GSP2XIVS.js";
+import "../chunk-3ISHDWO7.js";
+import "../chunk-NWLGL3IR.js";
 import "../chunk-KET4QQZB.js";
 import "../chunk-2N4OENGE.js";
-import "../chunk-UBZQE3JN.js";
+import "../chunk-UX3PQKVZ.js";
 import "../chunk-CQ553GZO.js";
 import "../chunk-ZP5E7YET.js";
-import "../chunk-X4EFYVCZ.js";
+import "../chunk-PC2R5RUJ.js";
 import "../chunk-GISMXMVL.js";
 import "../chunk-WWACIIBA.js";
 import "../chunk-TEFCFC4H.js";
-import "../chunk-2SNRCAC3.js";
+import "../chunk-KAJOS3BA.js";
 import {
   dispatchSkillsFromGit
-} from "../chunk-CAQPSFPT.js";
-import "../chunk-KBCPELBC.js";
-import "../chunk-LADPYELQ.js";
-import "../chunk-FZ5MBXEG.js";
-import "../chunk-ACOOG3UW.js";
-import "../chunk-UUO4K7NX.js";
+} from "../chunk-WWEVS7RO.js";
+import "../chunk-HEIMJJI4.js";
+import "../chunk-XX4OLNWQ.js";
+import "../chunk-YQSYCBAH.js";
+import "../chunk-Z4AWEIBY.js";
+import "../chunk-ONCPJGHY.js";
 import "../chunk-IOW3MW2K.js";
-import "../chunk-UXPWSV3B.js";
+import "../chunk-BYVT5LVO.js";
 import "../chunk-3ISINLYT.js";
 import "../chunk-CLI4K2CH.js";
-import "../chunk-KI6PHH6Q.js";
-import "../chunk-UJFNIWC7.js";
-import "../chunk-HRSE6IQD.js";
-import "../chunk-C7J55C6E.js";
+import "../chunk-HWTUUQOW.js";
+import "../chunk-527G27VI.js";
+import "../chunk-DE5U6KOL.js";
+import "../chunk-MGUPSE6D.js";
 import "../chunk-RC3OI5NK.js";
 import "../chunk-W6Y7ZW3Y.js";
 import "../chunk-27AJKSQY.js";
@@ -43,7 +43,7 @@ import "../chunk-HIK6OEUF.js";
 import {
   findConfigFile,
   loadConfig
-} from "../chunk-B7KMTHN3.js";
+} from "../chunk-GGKRA7A7.js";
 import "../chunk-ZCZD7FHU.js";
 import "../chunk-EBJQ6N4M.js";
 import {
@@ -56,7 +56,7 @@ import {
   readCheckState,
   shouldRunCheck,
   spawnBackgroundCheck
-} from "../chunk-GOHWCHCS.js";
+} from "../chunk-N24HCQA4.js";
 import {
   CLI_VERSION
 } from "../chunk-BM3PWGXQ.js";

package/dist/{check-phase-gate-MK364VXS.js → check-phase-gate-GNDOMS35.js}

@@ -1,12 +1,12 @@
 import {
   createCheckPhaseGateCommand,
   runCheckPhaseGate
-} from "./chunk-2SNRCAC3.js";
-import "./chunk-B7KMTHN3.js";
+} from "./chunk-KAJOS3BA.js";
+import "./chunk-GGKRA7A7.js";
 import "./chunk-ZCZD7FHU.js";
 import "./chunk-EBJQ6N4M.js";
 import "./chunk-3WGJMBKH.js";
-import "./chunk-GOHWCHCS.js";
+import "./chunk-N24HCQA4.js";
 import "./chunk-N55WOGN3.js";
 import "./chunk-BDGTZRZ6.js";
 import "./chunk-PP6ZRL5T.js";

package/dist/{chunk-VBNJAKUH.js → chunk-3ISHDWO7.js}

@@ -1,6 +1,6 @@
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Ok
 } from "./chunk-BDGTZRZ6.js";

package/dist/{chunk-UJFNIWC7.js → chunk-527G27VI.js}

@@ -1,6 +1,6 @@
 import {
   resolveProjectConfig
-} from "./chunk-HRSE6IQD.js";
+} from "./chunk-DE5U6KOL.js";
 import {
   resultToMcpResponse
 } from "./chunk-RC3OI5NK.js";
@@ -38,7 +38,7 @@ async function handleCheckDependencies(input) {
   const configResult = resolveProjectConfig(projectPath);
   if (!configResult.ok) return resultToMcpResponse(configResult);
   try {
-    const { validateDependencies, TypeScriptParser } = await import("./dist-JCFK263L.js");
+    const { validateDependencies, TypeScriptParser } = await import("./dist-LHINSVK4.js");
     const config = configResult.value;
     const rawLayers = Array.isArray(config.layers) ? config.layers : [];
     const layers = rawLayers.map((l) => ({

package/dist/{chunk-5XL5QJ5B.js → chunk-6WQCLCBO.js}

@@ -3,7 +3,7 @@ import {
 } from "./chunk-KET4QQZB.js";
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Err,
   Ok

package/dist/{chunk-UXPWSV3B.js → chunk-BYVT5LVO.js}

@@ -33,7 +33,7 @@ var createSelfReviewDefinition = {
 };
 async function handleCreateSelfReview(input) {
   try {
-    const { parseDiff, createSelfReview } = await import("./dist-JCFK263L.js");
+    const { parseDiff, createSelfReview } = await import("./dist-LHINSVK4.js");
     const parseResult = parseDiff(input.diff);
     if (!parseResult.ok) {
       return resultToMcpResponse(parseResult);
@@ -119,7 +119,7 @@ var analyzeDiffDefinition = {
 };
 async function handleAnalyzeDiff(input) {
   try {
-    const { parseDiff, analyzeDiff } = await import("./dist-JCFK263L.js");
+    const { parseDiff, analyzeDiff } = await import("./dist-LHINSVK4.js");
     const parseResult = parseDiff(input.diff);
     if (!parseResult.ok) {
       return resultToMcpResponse(parseResult);
@@ -195,7 +195,7 @@ var requestPeerReviewDefinition = {
 };
 async function handleRequestPeerReview(input) {
   try {
-    const { parseDiff, requestPeerReview } = await import("./dist-JCFK263L.js");
+    const { parseDiff, requestPeerReview } = await import("./dist-LHINSVK4.js");
     const parseResult = parseDiff(input.diff);
     if (!parseResult.ok) {
       return resultToMcpResponse(parseResult);

package/dist/{chunk-HRSE6IQD.js → chunk-DE5U6KOL.js}

@@ -1,6 +1,6 @@
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Err,
   Ok

package/dist/{chunk-B7KMTHN3.js → chunk-GGKRA7A7.js}

@@ -7,7 +7,7 @@ import {
 } from "./chunk-3WGJMBKH.js";
 import {
   init_dist as init_dist2
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Err,
   Ok

package/dist/{chunk-STESM73J.js → chunk-GSP2XIVS.js}

@@ -1,6 +1,6 @@
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Err,
   Ok

package/dist/{chunk-KBCPELBC.js → chunk-HEIMJJI4.js}

@@ -27,7 +27,7 @@ var runSecurityScanDefinition = {
 };
 async function handleRunSecurityScan(input) {
   try {
-    const core = await import("./dist-JCFK263L.js");
+    const core = await import("./dist-LHINSVK4.js");
     const projectRoot = sanitizePath(input.path);
     let configData = {};
     try {
@@ -114,7 +114,7 @@ var getSecurityTrendsDefinition = {
 };
 async function handleGetSecurityTrends(input) {
   try {
-    const core = await import("./dist-JCFK263L.js");
+    const core = await import("./dist-LHINSVK4.js");
     const projectRoot = sanitizePath(input.path);
     const manager = new core.SecurityTimelineManager(projectRoot);
     const trendOptions = {};

package/dist/{chunk-KI6PHH6Q.js → chunk-HWTUUQOW.js}

@@ -1,6 +1,6 @@
 import {
   resolveProjectConfig
-} from "./chunk-HRSE6IQD.js";
+} from "./chunk-DE5U6KOL.js";
 import {
   sanitizePath
 } from "./chunk-W6Y7ZW3Y.js";
@@ -49,7 +49,7 @@ async function handleValidateProject(input) {
   checks.config = "pass";
   const config = configResult.value;
   try {
-    const core = await import("./dist-JCFK263L.js");
+    const core = await import("./dist-LHINSVK4.js");
     if (typeof core.validateFileStructure === "function" && Array.isArray(config.conventions)) {
       const conventions = config.conventions;
       const structureResult = await core.validateFileStructure(projectPath, conventions);
@@ -68,7 +68,7 @@ async function handleValidateProject(input) {
   } catch {
   }
   try {
-    const core = await import("./dist-JCFK263L.js");
+    const core = await import("./dist-LHINSVK4.js");
     if (typeof core.validateAgentsMap === "function") {
       const agentsMapPath = path.join(projectPath, "AGENTS.md");
       const agentsResult = await core.validateAgentsMap(agentsMapPath);

package/dist/{chunk-2SNRCAC3.js → chunk-KAJOS3BA.js}

@@ -1,6 +1,6 @@
 import {
   resolveConfig
-} from "./chunk-B7KMTHN3.js";
+} from "./chunk-GGKRA7A7.js";
 import {
   logger
 } from "./chunk-EBJQ6N4M.js";
@@ -9,7 +9,7 @@ import {
 } from "./chunk-3WGJMBKH.js";
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Ok
 } from "./chunk-BDGTZRZ6.js";

package/dist/{chunk-C7J55C6E.js → chunk-MGUPSE6D.js}

@@ -7,10 +7,10 @@ import {
 import {
   findConfigFile,
   loadConfig
-} from "./chunk-B7KMTHN3.js";
+} from "./chunk-GGKRA7A7.js";
 import {
   init_dist
-} from "./chunk-GOHWCHCS.js";
+} from "./chunk-N24HCQA4.js";
 import {
   Ok
 } from "./chunk-BDGTZRZ6.js";
@@ -50,12 +50,12 @@ async function handleCheckDocs(input) {
     const projectPath = sanitizePath(input.path);
     const scope = input.scope ?? "coverage";
     if (scope === "integrity") {
-      const { validateKnowledgeMap } = await import("./dist-JCFK263L.js");
+      const { validateKnowledgeMap } = await import("./dist-LHINSVK4.js");
       const result2 = await validateKnowledgeMap(projectPath);
       return resultToMcpResponse(result2);
     }
     if (scope === "all") {
-      const { checkDocCoverage: checkDocCoverage2, validateKnowledgeMap } = await import("./dist-JCFK263L.js");
+      const { checkDocCoverage: checkDocCoverage2, validateKnowledgeMap } = await import("./dist-LHINSVK4.js");
       const domain2 = input.domain ?? "src";
       const { loadGraphStore: loadGraphStore2 } = await import("./graph-loader-HHXN5FLP.js");
       const store2 = await loadGraphStore2(projectPath);
@@ -94,7 +94,7 @@ async function handleCheckDocs(input) {
       }
       return resultToMcpResponse(Ok({ coverage, integrity }));
     }
-    const { checkDocCoverage } = await import("./dist-JCFK263L.js");
+    const { checkDocCoverage } = await import("./dist-LHINSVK4.js");
     const domain = input.domain ?? "src";
     const { loadGraphStore } = await import("./graph-loader-HHXN5FLP.js");
     const store = await loadGraphStore(projectPath);