fix(ci): 5 CI failures — Build V3, Static guards, ADR-112, witness verify, supply-chain audit

Stacked fixes to clear the remaining red CI checks on \`main\` that pre-dated
the #2412 lockfile-cascade fix. Each one verified locally before commit.

1. **Build V3 (ubuntu/macos/windows)** — \`graph-edge-writer.ts\` did a static
   \`await import('better-sqlite3')\` which forces tsc to resolve the package's
   types at build time. \`better-sqlite3\` is in \`optionalDependencies\` (so
   CI doesn't install it) — type resolution failed with TS2307 on every
   platform. Routed the module name through a variable so the static analyzer
   stops trying to resolve it; runtime presence is still gated by the
   existing try/catch.

2. **Static regression guards (#2267)** — \`scripts/smoke-workflows-yaml.mjs\`
   crashed with \`SyntaxError: The requested module 'js-yaml' does not provide
   an export named 'default'\` after js-yaml dropped the synthetic default
   from its package exports. Switched to \`import * as yaml from 'js-yaml'\`.

3. **Tool description discoverability (ADR-112)** — two MCP tools regressed
   the no-guidance baseline (count went from 0 → 2):
     - \`metaharness_bench\` had no Use-when / Prefer / Pair-with hint
     - \`testgen_tdd_repair\` had \"Use when …\" in a *later* string
       fragment of a concatenated description, which the audit regex
       (single-quoted-literal-only) misses
   Added explicit \"Use when …\" guidance to the first fragment of each.
   Audit now reports 340/340 tools with guidance.

4. **witness verify precondition smoke (#1880)** — Root \`package-lock.json\`
   drifted: \`agentdb@3.0.0-alpha.16\` in lock vs \`^3.0.0-alpha.17\` in
   package.json (same root cause as #2412 but for the npm lockfile, not the
   pnpm one fixed last release). Every job that runs \`npm ci\` was failing
   at install. Regenerated via \`npm install --package-lock-only\`. Verified:
   \`npm ci --dry-run --legacy-peer-deps\` now exits 0.

5. **Supply-chain audit (#2046)** — \`plugins/ruflo-graph-intelligence\` had
   1 HIGH/CRITICAL unaccepted direct-dep finding:
     - \`ws@8.20.1\` pulled by \`sublinear-time-solver\` (GHSA-96hv-2xvq-fx4p,
       memory exhaustion DoS) — bumped root override \`ws >=8.18.4\` →
       \`>=8.21.0\` and added the same override to the plugin's own
       package.json so it propagates through the plugin's own npm install
     - \`@claude-flow/cli\` peerDep surfaces every HIGH transitive in cli's
       own tree inside the plugin's audit report — added a tracked,
       expiring accepted-finding (cli ships its own audit pipeline on
       every publish; the plugin doesn't redistribute cli's tree).
   Audit now reports \`OK: no hard-fail findings\`.

Verified locally:
  - \`npm run build\` in v3/@claude-flow/cli: clean
  - \`node scripts/smoke-workflows-yaml.mjs\`: exit 0, 53 workflows parsed
  - \`node scripts/audit-tool-descriptions.mjs\`: 340/340 with guidance, OK
  - \`npm ci --dry-run --legacy-peer-deps\` at repo root: exit 0
  - \`node scripts/audit-supply-chain.mjs\`: OK, no hard-fail findings

Co-Authored-By: RuFlo <ruv@ruv.net>

.github/supply-chain/accepted-findings.json

@@ -39,6 +39,16 @@
       "trackingIssue": "https://github.com/ruvnet/ruflo/issues/2268",
       "expiresAt": "2026-09-02",
       "rationale": "devDependency-only + exploit requires `vitest --ui` which is local-dev only. Spec already requires ^4.1.0; only the standalone package-lock graph hasn't refreshed (workspace:* protocol limitation on raw `npm install`). Zero production / CI exposure. Re-audit and remove this entry the next time the browser's package-lock.json is regenerated."
+    },
+    {
+      "package": "plugins/ruflo-graph-intelligence",
+      "depName": "@claude-flow/cli",
+      "severity": "high",
+      "via": ["@claude-flow/embeddings", "@claude-flow/guidance"],
+      "tracking": "ruflo#2412 follow-up — @claude-flow/cli is declared as a peerDependency by ruflo-graph-intelligence; the plugin does not redistribute cli's dependency graph (the consumer installs @claude-flow/cli@latest separately). The audit script treats peerDependencies as direct deps, so it surfaces every HIGH transitive inside cli's own tree (xenova/transformers, agentdb's opentelemetry chain, onnxruntime-web) inside this plugin's report. Each of those transitives is independently triaged in its own package (see the v3/@claude-flow/browser entries above) and the cli already runs its own audit pipeline on every publish. Accepting this pair so the plugin's audit isn't transitively blocked on cli's chain.",
+      "trackingIssue": "https://github.com/ruvnet/ruflo/issues/2412",
+      "expiresAt": "2026-11-25",
+      "rationale": "Peer-dependency-only — the plugin does not redistribute cli's transitives. cli@3.14.2 ships its own audit on each publish. Re-audit and drop this entry once either (a) the peer is removed in favor of a slim interface package or (b) cli's tree no longer reports HIGH transitives."
     }
   ],
   "lockfile": [

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "claude-flow",
-  "version": "3.12.3",
+  "version": "3.14.2",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "claude-flow",
-      "version": "3.12.3",
+      "version": "3.14.2",
       "license": "MIT",
       "dependencies": {
         "@claude-flow/cli-core": "3.7.0-alpha.5",
@@ -15,7 +15,6 @@
         "@claude-flow/shared": "3.0.0-alpha.8",
         "@noble/ed25519": "2.3.0",
         "@ruvector/rabitq-wasm": "0.1.0",
-        "agentic-flow": "^2.0.14",
         "semver": "7.7.3",
         "zod": "3.25.76"
       },
@@ -46,7 +45,7 @@
         "@ruvector/router": "^0.1.30",
         "@ruvector/router-linux-x64-gnu": "^0.1.30",
         "@ruvector/sona": "^0.1.5",
-        "agentdb": "^3.0.0-alpha.16",
+        "agentdb": "^3.0.0-alpha.17",
         "agentic-flow": "^2.0.14"
       }
     },
@@ -3881,9 +3880,9 @@
       }
     },
     "node_modules/agentdb": {
-      "version": "3.0.0-alpha.16",
-      "resolved": "https://registry.npmjs.org/agentdb/-/agentdb-3.0.0-alpha.16.tgz",
-      "integrity": "sha512-hx5KildnxQ/mZaOaOKGHDe7OlDJmHoRV7sy5b4PpfjtHAE9dU0vzM7tpemZKPhnaRA89jxpKTutbnhux/JzgLg==",
+      "version": "3.0.0-alpha.17",
+      "resolved": "https://registry.npmjs.org/agentdb/-/agentdb-3.0.0-alpha.17.tgz",
+      "integrity": "sha512-LK8drZFQteNkff+ji52vCmIXyE3A3ipzabx+ubBcX0CRgF96GU8m36sWyOfqL7pWRfnYDgkEEEgZmagiNCaz/A==",
       "hasInstallScript": true,
       "license": "MIT",
       "dependencies": {

package.json

@@ -103,7 +103,7 @@
     "axios": ">=1.13.2",
     "fast-uri": ">=3.1.0",
     "vite": ">=6.4.6",
-    "ws": ">=8.18.4"
+    "ws": ">=8.21.0"
   },
   "devDependencies": {
     "@openai/codex": "^0.98.0",

plugins/ruflo-graph-intelligence/package-lock.json

@@ -1618,6 +1618,17 @@
         "node": ">=18"
       }
     },
+    "node_modules/@claude-flow/cli/node_modules/jose": {
+      "version": "5.10.0",
+      "resolved": "https://registry.npmjs.org/jose/-/jose-5.10.0.tgz",
+      "integrity": "sha512-s+3Al/p9g32Iq+oqXxkW//7jk2Vig6FF1CFqzVXoTUXt2qz89YWbL+OwS17NFYEvxC35n0FKeGO2LGYSxeM2Gg==",
+      "license": "MIT",
+      "optional": true,
+      "peer": true,
+      "funding": {
+        "url": "https://github.com/sponsors/panva"
+      }
+    },
     "node_modules/@claude-flow/cli/node_modules/media-typer": {
       "version": "1.1.0",
       "resolved": "https://registry.npmjs.org/media-typer/-/media-typer-1.1.0.tgz",
@@ -15304,9 +15315,9 @@
       "license": "ISC"
     },
     "node_modules/ws": {
-      "version": "8.20.1",
-      "resolved": "https://registry.npmjs.org/ws/-/ws-8.20.1.tgz",
-      "integrity": "sha512-It4dO0K5v//JtTXuPkfEOaI3uUN87iYPnqo/ZzqCoG3g8uhA66QUMs/SrM0YK7/NAu+r4LMh/9dq2A7k+rHs+w==",
+      "version": "8.21.0",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-8.21.0.tgz",
+      "integrity": "sha512-Vsp28b7DRcimFQvrqu2Wek3z1iYxDCWqHYB8Qsnk/S4RfaCQzPGPyBNuVjJV3cd6UiKtUtp6sNM77gWvzcCH+g==",
       "license": "MIT",
       "engines": {
         "node": ">=10.0.0"

plugins/ruflo-graph-intelligence/package.json

@@ -23,6 +23,9 @@
     "sublinear-time-solver": "^1.7.0",
     "zod": "^3.22.4"
   },
+  "overrides": {
+    "ws": ">=8.21.0"
+  },
   "peerDependencies": {
     "@claude-flow/cli": ">=3.5.0"
   },

scripts/smoke-workflows-yaml.mjs

@@ -20,7 +20,11 @@
 
 import { readFileSync, readdirSync } from 'node:fs';
 import { join } from 'node:path';
-import yaml from 'js-yaml';
+// js-yaml's package exports map removed the synthetic default in newer
+// versions when loaded as ESM, so `import yaml from 'js-yaml'` throws
+// `SyntaxError: does not provide an export named 'default'`. Use the
+// namespace import — works in every js-yaml release since 4.x.
+import * as yaml from 'js-yaml';
 
 const WORKFLOWS_DIR = '.github/workflows';
 

v3/@claude-flow/cli/src/mcp-tools/metaharness-tools.ts

@@ -443,7 +443,7 @@ export const metaharnessTools: MCPTool[] = [
   },
   {
     name: 'metaharness_bench',
-    description: 'ADR-153 supporting verb — create or verify bench suites used by metaharness_evolve --bench. Bench suites are JSON files of {input, expectedOutput, weight} tasks; scoring against a fixed corpus decouples evolution from flaky/slow/undersized `npm test`. Use --op create to scaffold from a repo, --op verify (cheap, ~5s) to gate suite changes in CI. Skipping bench suites is wrong when iterating on the same harness across commits because per-run noise drowns out champion-fitness deltas; bench gives you a stable baseline. ' + MCP_SUCCESS_SEMANTIC,
+    description: 'ADR-153 supporting verb — create or verify bench suites used by metaharness_evolve --bench. Bench suites are JSON files of {input, expectedOutput, weight} tasks; scoring against a fixed corpus decouples evolution from flaky/slow/undersized `npm test`. Use when iterating on the same harness across commits and `npm test` is too noisy/slow to drive evolution — use --op create to scaffold from a repo, --op verify (cheap, ~5s) to gate suite changes in CI. Native test runners are wrong here because per-run noise drowns out champion-fitness deltas; bench gives you a stable baseline. ' + MCP_SUCCESS_SEMANTIC,
     category: 'metaharness',
     inputSchema: {
       type: 'object',

v3/@claude-flow/cli/src/mcp-tools/testgen-tools.ts

@@ -124,7 +124,7 @@ export const testgenTools: MCPTool[] = [
   {
     name: 'testgen_tdd_repair',
     description:
-      'ADR-175-inspired Test-Driven Repair. Given a failing test, spawn a bounded headless `claude -p` (Read/Edit/Bash only, --max-budget-usd capped) ' +
+      'ADR-175-inspired Test-Driven Repair. Use when you have a failing test and want a verified, bounded-cost fix. Spawns a bounded headless `claude -p` (Read/Edit/Bash only, --max-budget-usd capped) ' +
       'that makes the test pass without modifying it. The test\'s exit code IS the fitness function — no LLM-as-judge. ' +
       'Use when you have a failing test and want a verified fix. Skipping this and asking an agent to "fix the bug" is wrong because (a) you lose the bounded-cost guarantee, ' +
       '(b) you lose the strict no-test-modification constraint, (c) you lose the verify-on-finish step. ' +

v3/@claude-flow/cli/src/memory/graph-edge-writer.ts

@@ -79,9 +79,17 @@ export async function getBridgeDb(customDbPath?: string, opts?: { createIfMissin
     // module, so platform-specific binaries are required). If unavailable,
     // return null and let callers degrade gracefully — same posture as the
     // prior sql.js version when sql.js failed to load.
+    //
+    // The module name is hidden behind a variable so the TypeScript compiler
+    // does not statically resolve and require the `better-sqlite3` types at
+    // build time — they would only be installed via optionalDependencies,
+    // which CI doesn't always install. This is the standard pattern for
+    // runtime-only optional native deps. The actual presence is gated by
+    // the try/catch below.
     let BetterSqlite3: any;
     try {
-      BetterSqlite3 = (await import('better-sqlite3')).default;
+      const mod: string = 'better-sqlite3';
+      BetterSqlite3 = (await import(mod)).default;
     } catch {
       return null;
     }