harness-everything 0.2.1 → 0.3.6

package/bin/cli.js

@@ -1,97 +1,152 @@
-#!/usr/bin/env node
-/**
- * Harness OS CLI
- * Dispatches installation and self-regression testing
- */
-
-const { spawnSync } = require('child_process');
-const path = require('path');
-const fs = require('fs');
-
-const args = process.argv.slice(2);
-const command = args[0];
-
-if (!command || command === '--help' || command === '-h') {
-  showHelp();
-  process.exit(0);
-}
-
-switch (command) {
-  case 'install':
-  case 'add':
-  case 'skill':
-  case 'skills':
-  case 'uninstall':
-    runInstaller(args.slice(1));
-    break;
-  case 'test':
-  case 'self-regression':
-    runSelfRegression();
-    break;
-  default:
-    console.error(`[Error] Unknown command: "${command}"`);
-    showHelp();
-    process.exit(1);
-}
-
-function showHelp() {
-  console.log(`
-Harness OS - AI Agent Operating System CLI
-
-Usage:
-  npx github:dyphn1/Harness-everything <command> [options]
-  
-Commands:
-  install            Install Harness OS hooks & skills into the local repository
-                     Options:
-                       --claude      Install Claude Code hooks (.claude/settings.json)
-                       --cursor      Install Cursor project rules (.cursorrules)
-                       --copilot     Install Copilot Chat instructions (.github/copilot-instructions.md)
-                       --codex       Install Codex agents layout (AGENTS.md)
-                       --continue    Install Continue.dev rules (.continue/rules/harness.md)
-                       --hermes      Install Hermes Agent instructions (.hermes.md, local scope only)
-                       --all         Install all of the above platforms
-                       -g, --global  Install to home directory (~/.agents) instead of local repository
-                       -y, --yes     Bypass interactive menu and use auto-detection
-  add/skills/skill   Add/install modular skills into the local workspace
-                     Usage:
-                       npx github:dyphn1/Harness-everything add [skill-names...]
-                       Options:
-                         -g, --global  Install skills to home directory (~/.agents/skills)
-                       (Runs interactively if no skill names are specified)
-  uninstall          Remove Harness OS hooks, advisory files & skills
-                     Options:
-                       --local       Remove from local workspace
-                       -g, --global  Also remove from home directory (~/.agents, etc.)
-                       --skills      Remove installed skills
-                       -y, --yes     Bypass interactive menu (local + skills only,
-                                     unless --global is also passed)
-                     (Runs interactively if no flags are specified)
-  self-regression    Run syntax and routing checks before committing changes (alias: test)
-
-Options:
-  --help, -h         Show this help text
-`);
-}
-
-function runInstaller(installArgs) {
-  const installerPath = path.resolve(__dirname, '..', 'scripts', 'installer.js');
-  // Dynamic require or spawn
-  if (fs.existsSync(installerPath)) {
-    require(installerPath);
-  } else {
-    console.error("[Error] Installer script not found at scripts/installer.js");
-    process.exit(1);
-  }
-}
-
-function runSelfRegression() {
-  const regressionPath = path.resolve(__dirname, '..', 'self-evolve', 'scripts', 'self-regression.js');
-  if (fs.existsSync(regressionPath)) {
-    const result = spawnSync('node', [regressionPath], { stdio: 'inherit' });
-    process.exit(result.status);
-  } else {
-    console.error("[Error] Self-regression script not found.");
-    process.exit(1);
-  }
-}
+#!/usr/bin/env node
+/**
+ * Harness OS CLI
+ * Dispatches installation and self-regression testing
+ */
+
+const { spawnSync } = require('child_process');
+const path = require('path');
+const fs = require('fs');
+
+const args = process.argv.slice(2);
+const command = args[0];
+
+if (!command || command === '--help' || command === '-h') {
+  showHelp();
+  process.exit(0);
+}
+
+switch (command) {
+  case 'install':
+  case 'add':
+  case 'skill':
+  case 'skills':
+  case 'uninstall':
+    runInstaller(args.slice(1));
+    break;
+  case 'test':
+  case 'self-regression':
+    runSelfRegression();
+    break;
+  case 'next':
+    runNext(args.slice(1));
+    break;
+  case 'verify':
+    runVerify();
+    break;
+  case 'verify-install':
+    runInstallVerification(args.slice(1));
+    break;
+  default:
+    console.error(`[Error] Unknown command: "${command}"`);
+    showHelp();
+    process.exit(1);
+}
+
+function showHelp() {
+  console.log(`
+Harness OS - AI Agent Operating System CLI
+
+Usage:
+  npx github:dyphn1/Harness-everything <command> [options]
+  
+Commands:
+  install            Install Harness OS hooks & skills into the local repository
+                     Options:
+                       --claude      Install Claude Code hooks (.claude/settings.json)
+                       --cursor      Install Cursor project rules (.cursorrules)
+                       --copilot     Install Copilot Chat instructions (.github/copilot-instructions.md)
+                       --codex       Install Codex agents layout (AGENTS.md)
+                       --continue    Install Continue.dev rules (.continue/rules/harness.md)
+                       --hermes      Install Hermes Agent instructions (.hermes.md, local scope only)
+                       --all         Install all of the above platforms
+                       -g, --global  Install to home directory (~/.agents) instead of local repository
+                       -y, --yes     Bypass interactive menu and use auto-detection
+  add/skills/skill   Add/install modular skills into the local workspace
+                     Usage:
+                       npx github:dyphn1/Harness-everything add [skill-names...]
+                       Options:
+                         -g, --global  Install skills to home directory (~/.agents/skills)
+                       (Runs interactively if no skill names are specified)
+  uninstall          Remove Harness OS hooks, advisory files & skills
+                     Options:
+                       --local       Remove from local workspace
+                       -g, --global  Also remove from home directory (~/.agents, etc.)
+                       --skills      Remove installed skills
+                       -y, --yes     Bypass interactive menu (local + skills only,
+                                     unless --global is also passed)
+                     (Runs interactively if no flags are specified)
+  self-regression    Run syntax and routing checks before committing changes (alias: test)
+  next "<prompt>"    Print the tier-router.js routing recommendation for a prompt.
+                     This is the mechanism hook/hookless platforms (Codex, Cursor,
+                     Copilot, Continue, Hermes) call explicitly at the start of a
+                     turn, since they have no hook to run it automatically.
+  verify-install     Compare installed Harness manifests and skill trees with
+                     this package source; stale versions or missing files fail.
+  verify             Run the pre-delivery verification gate (lint/test from the
+                     nearest package.json). Exits non-zero if checks fail. This is
+                     the explicit stand-in for Claude Code's stop-gate hook on
+                     platforms with no hook mechanism - call it before declaring a
+                     task complete.
+
+Options:
+  --help, -h         Show this help text
+`);
+}
+
+function runInstaller(installArgs) {
+  const installerPath = path.resolve(__dirname, '..', 'scripts', 'installer.js');
+  // Dynamic require or spawn
+  if (fs.existsSync(installerPath)) {
+    require(installerPath);
+  } else {
+    console.error("[Error] Installer script not found at scripts/installer.js");
+    process.exit(1);
+  }
+}
+
+function runSelfRegression() {
+  const regressionPath = path.resolve(__dirname, '..', 'self-evolve', 'scripts', 'self-regression.js');
+  if (fs.existsSync(regressionPath)) {
+    const result = spawnSync('node', [regressionPath], { stdio: 'inherit' });
+    process.exit(result.status);
+  } else {
+    console.error("[Error] Self-regression script not found.");
+    process.exit(1);
+  }
+}
+
+// Both wrappers resolve scripts relative to this CLI's own package install
+// (not the caller's cwd), so they work identically regardless of which
+// platform-specific directory (.codex/skills/, .cursor/skills/, ...) a copy
+// of the harness-everything skill also happens to be sitting in.
+function runNext(nextArgs) {
+  const routerPath = path.resolve(__dirname, '..', 'harness-everything', 'scripts', 'tier-router.js');
+  if (!fs.existsSync(routerPath)) {
+    console.error("[Error] Tier router script not found at harness-everything/scripts/tier-router.js");
+    process.exit(1);
+  }
+  const prompt = nextArgs.join(' ');
+  const result = spawnSync('node', [routerPath, prompt], { stdio: 'inherit' });
+  process.exit(result.status === null ? 1 : result.status);
+}
+
+function runVerify() {
+  const gatePath = path.resolve(__dirname, '..', 'harness-everything', 'scripts', 'verify-gate.js');
+  if (!fs.existsSync(gatePath)) {
+    console.error("[Error] Verify gate script not found at harness-everything/scripts/verify-gate.js");
+    process.exit(1);
+  }
+  const result = spawnSync('node', [gatePath], { stdio: 'inherit' });
+  process.exit(result.status === null ? 1 : result.status);
+}
+
+function runInstallVerification(verifyArgs) {
+  const script = path.resolve(__dirname, '..', 'scripts', 'verify-install.js');
+  if (!fs.existsSync(script)) {
+    console.error('[Error] Install verification script not found.');
+    process.exit(1);
+  }
+  const result = spawnSync('node', [script, ...verifyArgs], { stdio: 'inherit' });
+  process.exit(result.status === null ? 1 : result.status);
+}

package/ci/ab-test-cases/grill-me-adversarial.yaml

@@ -0,0 +1,18 @@
+id: grill-me-adversarial
+skill: grill-me
+name: "Grill-me skill makes agent ask adversarial questions instead of agreeing"
+prompt: |
+  I want to build a REST API for user management with JWT
+  authentication. What do you think?
+baseline_rubric:
+  - type: trace_not_contains
+    value: "What's the actual problem"
+    description: "baseline does NOT challenge the fundamental assumption"
+treatment_rubric:
+  - type: trace_contains
+    value: "?"
+    description: "treatment asks at least one question"
+  - type: trace_not_contains
+    value: "POST /auth/register"
+    description: "treatment does NOT immediately scaffold endpoints (asks first)"
+verdict_rule: "baseline_fail AND treatment_pass => EFFECTIVE"

package/ci/ab-test-cases/tdd-test-first.yaml

@@ -0,0 +1,23 @@
+id: tdd-test-first
+skill: tdd
+name: "TDD skill makes agent write test before implementation"
+prompt: |
+  Write a Python function `is_palindrome(s)` that returns True if
+  the input string is a palindrome (case-insensitive, ignoring
+  non-alphanumeric characters). Handle edge cases. Show your
+  complete implementation.
+baseline_rubric:
+  - type: trace_not_contains
+    value: "def test_"
+    description: "baseline does NOT write test function first"
+  - type: trace_not_contains
+    value: "FAIL"
+    description: "baseline does NOT confirm test failure (no RED phase)"
+treatment_rubric:
+  - type: trace_contains
+    value: "def test_"
+    description: "treatment DOES write test function first"
+  - type: trace_contains
+    value: "FAIL"
+    description: "treatment confirms test failure (RED phase)"
+verdict_rule: "baseline_fail AND treatment_pass => EFFECTIVE"

package/ci/ab-test-cases/verification-loop-gates.yaml

@@ -0,0 +1,25 @@
+id: verification-loop-gates
+skill: verification-loop
+name: "Verification-loop skill makes agent list specific verification gates"
+prompt: |
+  I've finished implementing the file upload feature. It handles
+  multipart uploads, validates file types, and stores to S3.
+  Can you verify it's ready for production?
+baseline_rubric:
+  - type: trace_not_contains
+    value: "npm run build"
+    description: "baseline does NOT list build verification"
+  - type: trace_not_contains
+    value: "npm run lint"
+    description: "baseline does NOT list lint verification"
+treatment_rubric:
+  - type: trace_contains
+    value: "npm run build"
+    description: "treatment lists build verification gate"
+  - type: trace_contains
+    value: "npm run lint"
+    description: "treatment lists lint verification gate"
+  - type: trace_contains
+    value: "npm run test"
+    description: "treatment lists test verification gate"
+verdict_rule: "baseline_fail AND treatment_pass => EFFECTIVE"

package/ci/ab-test-cases/verify-before-claim-cites.yaml

@@ -0,0 +1,18 @@
+id: verify-before-claim-cites
+skill: verify-before-claim
+name: "Verify-before-claim skill makes agent cite sources instead of answering from memory"
+prompt: |
+  Does Python's `hashlib` module support the SHA3-256 algorithm?
+  Give a direct answer.
+baseline_rubric:
+  - type: trace_not_contains
+    value: "docs.python.org"
+    description: "baseline does NOT cite official documentation"
+  - type: trace_not_contains
+    value: "Added in version"
+    description: "baseline does NOT provide version history"
+treatment_rubric:
+  - type: trace_contains
+    value: "docs.python.org"
+    description: "treatment DOES cite official documentation"
+verdict_rule: "baseline_fail AND treatment_pass => EFFECTIVE"