@clear-capabilities/agentic-security-scanner 0.139.1 → 0.141.0

package/CHANGELOG.md

@@ -9,6 +9,227 @@
 > make the history less accurate, not more.
 
 
+
+## 0.141.0 — Six new instruments, and the seven live bugs they found
+
+The previous release fixed five bugs found by measuring. This one builds the
+instruments that do the measuring for the surfaces that had none — SCA, secrets,
+IaC, prompt injection, remediation, and the IDE/MCP surfaces — and then fixes
+what they reported. Every number below is reproducible with a command in
+`bench/*/README.md`.
+
+None of the improvements came from tuning. They came from files that were never
+read, patterns that could never fire, and versions that were silently rewritten.
+
+### SCA never read a transitive dependency tree on any real project
+
+`readTree` skipped **any file over 500 KB before deciding what kind of file it
+was.** npm/cli's `package-lock.json` is 666 KB, next.js's `pnpm-lock.yaml` is
+910 KB, magento2's `composer.lock` is 501 KB. On every project big enough for
+supply-chain risk to matter, the lockfile was dropped and SCA fell back to
+whatever exact versions appeared in `package.json` — direct dependencies only,
+while the headline claim of the feature is transitive reachability.
+
+Manifests now have their own, much larger cap. The 500 KB cap on **code** files
+is unchanged: that one protects the analysis path and was never the problem.
+
+Two more admission gaps in the same area: `go.sum` was never admitted though
+`_parseGoSum` and its dispatch entry had always existed, and only the exact
+basename `requirements.txt` was matched — so `requirements/dev.txt`, which is
+what pallets/flask ships, scored 0 of 11.
+
+**Measured effect (`bench/sca-replay`, 13 real repos, 7 ecosystems, labels from
+the advisory database via readers that share no code with the engine): version
+recall 10.89% → 77.92% at 100% precision, held-out 84.29%.**
+
+### Go dependency versions were rewritten into versions that do not exist
+
+`v0.0.0-20210903162142-ad29c8ab022f` became `0.0.0` in three separate places —
+not a shorter version but a different one, collapsing every pseudo-versioned
+module in a tree onto a single key. **Go went from 5.28% to 100%** once versions
+survived intact.
+
+This also corrupted the emitted SBOM, which is the worse half: an SBOM saying
+`golang.org/x/net@0.0.0` is wrong in a document other people are supposed to
+rely on.
+
+### The typosquat detector reported 166 findings, none of them typosquats
+
+Across 13 real repositories, at critical and high severity: `ms ~ ws`,
+`acorn ~ cors`, `ajv ~ ava`, `six ~ tox`, `arg ~ yargs`, `bail ~ babel`. Every
+one is a legitimate, popular package; `ms` is a top-50 npm package.
+
+Absolute edit distance is meaningless on short names — two edits on a
+four-character name changes half of it, and every two-character package is one
+edit from every other. Now Damerau-Levenshtein, so a transposition (`lodahs` for
+`lodash`, the commonest real typo) costs 1 rather than 2, gated on
+`distance / min(len) ≤ 0.25`.
+
+### The VS Code extension had never been able to find the scanner
+
+It looked for the bundle under a hardcoded
+`…/agentic-security/0.1.0/scanner/dist/…`, and Claude Code caches a plugin under
+its **plugin** version — 0.128.2, 0.136.9, 0.139.1, never 0.1.0. `0.1.0` was the
+extension's own version pasted into the wrong path, so the fallback could not
+resolve on any install and every user got *"scanner not found."*
+
+Nothing tested it, because the function read `vscode.workspace` and could not be
+imported outside a VS Code host. The resolver is now a pure function that
+discovers the version instead of hardcoding one, prefers `CLAUDE_PLUGIN_ROOT`,
+and is covered by 15 tests plus a CI job that fails on a stale committed bundle.
+
+### Three IaC formats had no rules at all
+
+CloudFormation, Bicep and Helm chart values scored **0** against every control
+tested — not weak rules, no rules, and for CloudFormation no file admission
+either, since a template is a `.yaml` that no path predicate recognises.
+
+`bench/iac-coverage` scores **verdict flip**: a control counts only when the
+misconfigured variant fires *and* the hardened one stays silent. That caught
+something a recall-only bench never would — `FROM ubuntu@sha256:…`, the most
+tightly pinned form a Dockerfile can use, was reported as *"ubuntu:latest
+(floating tag)"* because the digest was matched but not captured. A false
+positive on the hardened configuration tells the people who did the right thing
+that they did the wrong one.
+
+**Coverage 8/14 → 23/26**, with the three still open marked as needing semantic
+analysis rather than another pattern.
+
+### Credential patterns that could never fire
+
+`CRED_PREFILTER` is a whole-file gate: `scanCredentials` returns early unless
+that one regex matches, so a pattern whose trigger token is missing is dead code
+however correct it is. The generic "Password in URL" rule was in exactly that
+state.
+
+Also absent entirely: `postgres://user:pass@host/db` and `mongodb+srv://…` —
+among the commonest real leaks there are — plus GitLab, DigitalOcean, Azure
+Storage, Supabase and HubSpot tokens. Only `jdbc:` was covered.
+
+**Format coverage 60% → 92.11%, with correct silence at 28/28** on a hard
+negative set of lockfile integrity fields, git SHAs, content digests, Terraform
+state ids and a security rule file that defines key formats.
+
+### Prompt-injection patterns that were correct and far too literal
+
+The override rule required the object noun to be one of seven words, so *"Forget
+all previous **tasks**"* and *"Ignore all preceding **orders**"* missed.
+Exfiltration could not match *"show me all your prompt texts"*.
+
+**Recall 6.08% → 18.25% against `deepset/prompt-injections` (Apache-2.0, 263
+injections and 399 legitimate prompts), precision and correct-silence at 100%
+throughout.** Reported per technique, because an aggregate hides which one is
+weak: role-play 100%, exfiltration 60%, override 41.86%.
+
+And the engine caught its own author — the widened pattern was flagged by this
+project's own ReDoS detector on the self-scan gate, twice, before it was
+rewritten as a flat alternation that scores identically.
+
+## Measured, published, and deliberately not "fixed"
+
+- **Fix synthesis produces a patch for 0 of 6** real true positives on
+  third-party code (`bench/fix-correctness`, scored against the upstream fix
+  commit). The deterministic synthesizer has two rules, both JS/Python; the real
+  population is injection and authorization across seven languages. Widening it
+  to guess at an authorization check would produce patches that pass
+  verification while changing behaviour.
+- **German prompt injection scores 2.30% against English's 28.57%.** Patching it
+  because *this corpus* is German would be fitting to the benchmark.
+- **Datadog, Vercel and Algolia keys stay undetected.** They are a bare run of
+  hex; a pattern for "32 hex characters" would fire on every content digest and
+  checksum in existence.
+- **`CODEGEN` produces nothing on any of the five advisories in its own header.**
+  Measured dead, and left for a deliberate retire-or-fix decision rather than
+  removed silently.
+
+## Also
+
+- The advisory miner **was not paginating** — the API ignores `&page=N` — so the
+  evaluation population was capped at ~100 advisories per ecosystem by
+  construction. Fixed via the `Link` cursor: **315 → 1004 entries**, Ruby 32 →
+  250, Kotlin 0 → 2.
+- `File.join(<root>, …, <variable>)` reaching a filesystem call with no traversal
+  guard is now detected for Ruby (CWE-22), the dominant real-world shape the
+  existing rule could not reach.
+- A guard-shaped method **name** on a declaration line (`def check_static_cache(`)
+  no longer counts as a containment guard, which was silently dropping path
+  findings inside any method called `check*`/`validate*`/`ensure*`.
+- MCP is now smoke-tested end to end over stdio through the shipped binary, with
+  both write tools asserted to refuse out-of-tree paths in both directions.
+
+## 0.140.0 — Five shipped bugs, found by measuring instead of reading
+
+Every fix here is a defect that was live in 0.139.1. None came from the feature
+backlog; all five came from measuring the engine against real code and taking
+failing signals seriously instead of explaining them away.
+
+### `scan --format sarif` produced INVALID SARIF in CI
+
+The CLI dispatches every command as `process.exit(await cmdX(args))`, and
+`process.exit()` does not flush an asynchronous stdout. stdout is asynchronous
+exactly when it is a pipe — every `> results.sarif`, `| jq`, and CI capture — so
+output was discarded at the 64 KiB pipe boundary, mid-token, with a normal exit
+status. On one directory that was 65,536 bytes emitted of 390,177: roughly 83%
+of the document silently dropped.
+
+**If you upload SARIF to code scanning, this affected you on any project large
+enough to matter.** A TTY and a file both flush synchronously, which is why it
+looked fine by hand and broke in automation.
+
+### LLM01 — Prompt Injection — could never fail
+
+Detectors emit the finding family as `<family>-<rule-slug>`
+(`prompt-injection-http-user-input-in-llm-`), while the compliance evaluator
+resolved `family:prompt-injection` as an exact key. It matched nothing, so the
+control reported as evidenced no matter what the scan found. The first control
+of the OWASP LLM Top 10 was structurally incapable of failing, along with ASVS
+V5.1 and NIST AI 600-1 MG-3.2-005.
+
+Two further compliance defects in the same matching code: an empty family
+bucket rendered as `✓ no open critical/high findings`, so two controls (ASVS
+V7.1, NIST Privacy CT.DP-P1) read `present` on every scan of every project; and
+a first attempt at fixing that wrongly declared four live families
+unevidenceable, degrading 15 working controls. Both directions are now gated —
+a control with no possible evidence cannot read `present`, and a family with a
+producer cannot be declared a gap.
+
+### `--deterministic` did not produce deterministic output
+
+Four of ten emitted formats differed run to run: CycloneDX/SPDX document ids and
+a CycloneDX bom-ref fallback from `crypto.randomUUID()`, a PoC marker from
+`Math.random()`, and per-file wall-clock timings (which also determined the
+sort ORDER, so blanking the values alone would not have been enough).
+
+**An attestation over an SBOM was therefore unverifiable** — the point of
+signing an artifact is that someone can regenerate and compare it.
+
+### 62% of concurrency findings on Go were false positives
+
+The lock guard matched a bare receiver (`defer mu.Unlock()`) but not a qualified
+one (`defer s.mu.Unlock()`), which is how a mutex held as a struct field is
+always written. The acquire pattern always matched the qualified form, so the
+two halves of the rule had disagreed since it was written — and the most
+idiomatic CORRECT code was the most likely to be reported. Measured: 170 of 273
+findings on a Go sample were false positives. Those findings also carried no
+CWE, so they were invisible to every CWE-keyed report.
+
+### New gates
+
+- `test/stdout-flush.test.js` — spawns the real CLI through a real pipe
+- `test/format-determinism.test.js` — every emitted format, byte-compared
+- `test/compliance-mapping-liveness.test.js` — both vacuous-pass directions
+- `test/concurrency-cwe.test.js` — per-lock guard discrimination
+- `bench/family-producers/OBSERVED.json` — 213 families observed across 331
+  real scan roots, recorded explicitly as a LOWER BOUND
+
+### Scope
+
+The world-class-harness PRD is **partially delivered**. F10.5 (determinism as a
+published property) is complete. F10.2 is half done — the enforcement half
+landed; the measurement half needs detectors to declare their families, because
+this release proved no textual search can enumerate them. Roughly 35 PRD items
+remain, several blocked on design decisions rather than implementation.
+
 ## 0.139.1 — The same scanner, published with provenance
 
 **The shipped artifact is functionally identical to 0.139.0.** Only two commits

package/bin/agentic-security.js

@@ -9,6 +9,35 @@ const __require = createRequire(import.meta.url);
 const PKG_VERSION = __require('../package.json').version;
 import { signLastScan as _signLastScan, verifyLastScan as _verifyLastScanShared } from '../src/posture/integrity.js';
 import { runScan } from '../src/runScan.js';
+
+// Every command is dispatched as `process.exit(await cmdX(args))`, and
+// process.exit() does NOT flush an asynchronous stdout. stdout is asynchronous
+// whenever it is a PIPE — which is every `> file`, `| jq`, and CI capture — so
+// anything still buffered when the process exits is discarded at the pipe
+// boundary: 64 KiB on macOS and Linux.
+//
+// That silently truncated `scan --format sarif` mid-token for any project
+// large enough to matter, i.e. the primary CI integration path, while still
+// exiting with a normal status. The consumer sees a JSON parse error with no
+// connection to its cause, or ingests a partial finding set.
+//
+// fs.writeSync(1, …) hands the bytes to the OS before returning, so a later
+// exit cannot lose them. A non-blocking pipe can still short-write or raise
+// EAGAIN, hence the loop — a partial write that is not retried is the same
+// truncation bug wearing a different hat.
+function writeStdout(s) {
+  const buf = Buffer.from(String(s), 'utf8');
+  let off = 0;
+  while (off < buf.length) {
+    try {
+      off += fs.writeSync(1, buf, off, buf.length - off);
+    } catch (e) {
+      if (e.code === 'EAGAIN') continue;   // pipe full; the reader will drain it
+      if (e.code === 'EPIPE') return;      // reader closed (`| head`) — not our error
+      throw e;
+    }
+  }
+}
 import { toJSON, toMarkdown, toSARIF, toSTIX, toCSV, toJUnit, toCLI, toCLIByProfile, toShipVerdict, toProTable, toHTML, toSummary, toVex, exitCodeFor, normalizeFindings } from '../src/report/index.js';
 import { toCycloneDX, toSPDX } from '../src/posture/sbom.js';
 import { toPBOM } from '../src/sast/pipeline.js';
@@ -679,7 +708,7 @@ async function cmdScan(args) {
   }
 
   if (output) await fsp.writeFile(output, body);
-  else process.stdout.write(body + '\n');
+  else writeStdout(body + '\n');
 
   // Persist last scan for /security-fix and /security-report
   const { isSafeStateDir: _isSafeStateDir, stateWritesEnabled: _writesOnScan } = await import('../src/posture/state-dir.js');
@@ -1251,7 +1280,7 @@ async function cmdHarness(args) {
     body += `\n\nHarnesses discovered: ${present.join(', ')}${includeHome ? ' (project + ~/)' : ' (project only)'}\n`;
   }
   if (args.flags.output) await fsp.writeFile(args.flags.output, body);
-  else process.stdout.write(body + '\n');
+  else writeStdout(body + '\n');
   return exitCodeFor(scan);
 }
 
@@ -1746,7 +1775,7 @@ async function cmdCompliance(args) {
 
   if (args.flags.list) {
     const fws = listFrameworks(scanRoot);
-    if (fmt === 'json') { console.log(JSON.stringify(fws, null, 2)); return 0; }
+    if (fmt === 'json') { writeStdout(JSON.stringify(fws, null, 2) + '\n'); return 0; }
     for (const f of fws) console.log(`  ${f.id.padEnd(20)} ${f.name}  [${f.source}]`);
     return 0;
   }
@@ -1780,7 +1809,7 @@ async function cmdCompliance(args) {
 
   const gapsOnly = !!args.flags.gap;
   if (fmt === 'json') {
-    console.log(JSON.stringify(gapsOnly ? { ...r, controls: r.controls.filter(c => c.bucket === 'gap') } : r, null, 2));
+    writeStdout(JSON.stringify(gapsOnly ? { ...r, controls: r.controls.filter(c => c.bucket === 'gap') } : r, null, 2) + '\n');
   } else if (fmt === 'md') {
     console.log(fs.readFileSync(statePath(scanRoot, 'privacy-framework.md'), 'utf8'));
   } else {
@@ -2018,7 +2047,7 @@ async function cmdFix(args) {
 
   // Default mode: print the canonical template (back-compat — security-fixer subagent applies it).
   if (!isPreview && !isApply) {
-    console.log(JSON.stringify(f, null, 2));
+    writeStdout(JSON.stringify(f, null, 2) + '\n');
     if (f.fix?.code) { console.log('\n--- suggested patch ---\n'); console.log(f.fix.code); }
     console.log('\nUse --preview to see a diff, or --apply to apply directly.');
     return 0;
@@ -2247,7 +2276,7 @@ async function main() {
         const { analyzeTranscript, formatCacheReport } = await import('../src/posture/cache-economics.js');
         const projectDir = path.resolve(args.flags.root || process.env.CLAUDE_PROJECT_DIR || process.cwd());
         const result = analyzeTranscript({ transcriptPath: args.flags.transcript, projectDir });
-        if (args.flags.json) console.log(JSON.stringify(result, null, 2));
+        if (args.flags.json) writeStdout(JSON.stringify(result, null, 2) + '\n');
         else console.log(formatCacheReport(result));
         process.exit(0);
       }
@@ -2291,7 +2320,7 @@ async function main() {
         });
         if (args.flags.json) {
           // Stringify Set/etc. safely.
-          console.log(JSON.stringify(r, null, 2));
+          writeStdout(JSON.stringify(r, null, 2) + '\n');
         } else if (!r.ok) {
           console.error(`cve-watch: ${r.reason || 'failed'}`);
         }
@@ -2307,7 +2336,7 @@ async function main() {
         const headRef = args.flags.head || args.flags.h || 'HEAD';
         if (!baseRef) { console.error('pr-delta: --base <ref> is required'); process.exit(2); }
         const delta = await computePrDelta(path.resolve(root), { baseRef, headRef });
-        if (args.flags.json) console.log(JSON.stringify(delta, null, 2));
+        if (args.flags.json) writeStdout(JSON.stringify(delta, null, 2) + '\n');
         else console.log(renderPrDeltaText(delta));
         // Exit non-zero if any critical/high introduced (useful as CI gate).
         const i = delta.summary?.introduced || {};
@@ -2359,7 +2388,7 @@ async function main() {
         const repo = args.flags.repo;
         if (!repo) { console.error('leaderboard-row: --repo <owner/name> is required'); process.exit(2); }
         const row = leaderboardRowFor({ scanRoot: path.resolve(root), repo });
-        console.log(JSON.stringify(row, null, 2));
+        writeStdout(JSON.stringify(row, null, 2) + '\n');
         process.exit(0);
       }
       case 'history': {
@@ -2372,7 +2401,7 @@ async function main() {
           since:    args.flags.since    || '6.months',
           interval: args.flags.interval || '1.month',
         });
-        if (args.flags.json) console.log(JSON.stringify(r, null, 2));
+        if (args.flags.json) writeStdout(JSON.stringify(r, null, 2) + '\n');
         else if (r.error) console.error(`history: ${r.error}`);
         else {
           console.log(`Scanned ${r.refs.length} refs.`);
@@ -2408,7 +2437,7 @@ async function main() {
           ? (Array.isArray(args.flags.remove) ? args.flags.remove : [args.flags.remove])
           : [];
         const r = await runWhatIf(path.resolve(root), { overlays, remove });
-        if (args.flags.json) console.log(JSON.stringify(r, null, 2));
+        if (args.flags.json) writeStdout(JSON.stringify(r, null, 2) + '\n');
         else {
           console.log(`baseline: ${r.baselineFindings} findings`);
           console.log(`what-if:  ${r.whatIfFindings} findings (delta ${r.delta >= 0 ? '+' : ''}${r.delta})`);

package/dist/113.index.js

@@ -10,7 +10,7 @@ export const modules = {
 /* harmony export */   fixDurationReport: () => (/* binding */ fixDurationReport),
 /* harmony export */   renderFixDurationSummary: () => (/* binding */ renderFixDurationSummary)
 /* harmony export */ });
-/* unused harmony exports FIX_STAGES, loadFixAttempts, bucketOf, summarizeFixDurations, _internals */
+/* unused harmony exports FIX_STAGES, loadFixAttempts, bucketOf, summarizeFixDurations, _internals, summarizeFixAxes, renderFixAxes */
 /* harmony import */ var node_fs__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(3024);
 /* harmony import */ var node_path__WEBPACK_IMPORTED_MODULE_1__ = __webpack_require__(6760);
 /* harmony import */ var _state_dir_js__WEBPACK_IMPORTED_MODULE_2__ = __webpack_require__(1174);
@@ -213,6 +213,82 @@ function renderFixDurationSummary(sum) {
 const _internals = { _dist, _pct, RELIABLE_N };
 
 
+// ── PRD F6.1 — score fixes on THREE AXES, not one ──────────────────────────
+//
+// The three axes the PRD names:
+//   (a) does the finding disappear      — the rescan leg
+//   (b) does the project's own suite pass — the tests leg
+//   (c) does an independent verifier agree — the PoC re-check leg
+//
+// All three were already computed by verifyFixCore and then collapsed into one
+// boolean, which is the problem: **(a) alone is satisfiable by deleting code.**
+// A patch that removes the vulnerable function passes the rescan, has nothing
+// left to fail, and — on a project with no detectable test suite — reaches
+// ok:true having proven only that the detector went quiet.
+//
+// Reporting the axes separately makes that visible. `aOnly` is the number that
+// matters most and the one nobody was publishing: attempts that satisfied ONLY
+// the disappearance axis. A high aOnly with a high headline is the shape of a
+// remediation feature that is deleting code and calling it a fix.
+function summarizeFixAxes(attempts) {
+  const list = Array.isArray(attempts) ? attempts.filter(Boolean) : [];
+  const d = list.length;
+
+  const rate = (pred) => ({ n: list.filter(pred).length, d });
+
+  // Each axis is judged INDEPENDENTLY of the overall verdict, so a leg that
+  // passed inside a failed attempt still counts for its own axis. Reading them
+  // off `ok` would make the three axes three copies of the same number.
+  const findingDisappeared = rate((a) => a.rescanOk === true || (a.ok === true && a.rescanOk !== false));
+  const testsStillPass = rate((a) => a.testsRan === true && a.testsOk !== false);
+  const verifierAgrees = rate((a) => a.pocOk === true);
+
+  const satisfiesAll = rate((a) =>
+    (a.rescanOk === true || (a.ok === true && a.rescanOk !== false))
+    && a.testsRan === true && a.testsOk !== false
+    && a.pocOk === true);
+
+  // The honesty number: disappearance WITHOUT either corroborating axis.
+  const aOnly = rate((a) => {
+    const disappeared = a.rescanOk === true || (a.ok === true && a.rescanOk !== false);
+    const corroborated = (a.testsRan === true && a.testsOk !== false) || a.pocOk === true;
+    return disappeared && !corroborated;
+  });
+
+  return {
+    total: d,
+    findingDisappeared,
+    testsStillPass,
+    verifierAgrees,
+    satisfiesAll,
+    aOnly,
+    meaning:
+      'findingDisappeared = the detector went quiet; testsStillPass = the project suite ran AND passed; '
+      + 'verifierAgrees = an independent PoC re-check confirmed the hole is shut. '
+      + 'aOnly counts attempts that satisfied ONLY disappearance — the shape a code-deleting "fix" produces.',
+    caveat: d === 0
+      ? 'no attempts recorded; every rate is 0/0 and means nothing'
+      : 'rates carry {n,d}; a small d is indicative, not settled',
+  };
+}
+
+/** Markdown for a report. Denominators always attached. */
+function renderFixAxes(sum) {
+  if (!sum || !sum.total) return '_No fix attempts recorded._\n';
+  const row = (label, r, note) => `| ${label} | ${r.n}/${r.d} | ${note} |`;
+  return [
+    '| Axis | Rate | Meaning |',
+    '|---|---|---|',
+    row('(a) finding disappeared', sum.findingDisappeared, 'the detector went quiet'),
+    row('(b) project tests pass', sum.testsStillPass, 'the suite RAN and passed'),
+    row('(c) verifier agrees', sum.verifierAgrees, 'an independent PoC re-check confirmed it'),
+    row('all three', sum.satisfiesAll, 'the only row that means "fixed"'),
+    row('(a) ALONE', sum.aOnly, 'satisfiable by deleting code — watch this number'),
+    '',
+  ].join('\n');
+}
+
+
 /***/ }),
 
 /***/ 4113:
@@ -415,8 +491,8 @@ var external_node_child_process_ = __webpack_require__(1421);
 var external_node_fs_ = __webpack_require__(3024);
 // EXTERNAL MODULE: external "node:path"
 var external_node_path_ = __webpack_require__(6760);
-// EXTERNAL MODULE: ./src/engine.js + 203 modules
-var engine = __webpack_require__(6622);
+// EXTERNAL MODULE: ./src/engine.js + 200 modules
+var engine = __webpack_require__(3474);
 ;// CONCATENATED MODULE: ./src/posture/fix-honesty-gate.js
 // Deterministic honesty gates on fix / finding output (#7).
 //

package/dist/178.index.js

@@ -13,7 +13,7 @@ export const modules = {
 /* harmony import */ var node_child_process__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(1421);
 /* harmony import */ var node_fs__WEBPACK_IMPORTED_MODULE_1__ = __webpack_require__(3024);
 /* harmony import */ var node_path__WEBPACK_IMPORTED_MODULE_2__ = __webpack_require__(6760);
-/* harmony import */ var _engine_js__WEBPACK_IMPORTED_MODULE_3__ = __webpack_require__(6622);
+/* harmony import */ var _engine_js__WEBPACK_IMPORTED_MODULE_3__ = __webpack_require__(3474);
 // Time-travel + counterfactual scanning (v0.68).
 //
 // Two new modes that exploit the pure-input shape of runFullScan:

package/dist/238.index.js

@@ -10,7 +10,7 @@ export const modules = {
 /* harmony export */   fixDurationReport: () => (/* binding */ fixDurationReport),
 /* harmony export */   renderFixDurationSummary: () => (/* binding */ renderFixDurationSummary)
 /* harmony export */ });
-/* unused harmony exports FIX_STAGES, loadFixAttempts, bucketOf, summarizeFixDurations, _internals */
+/* unused harmony exports FIX_STAGES, loadFixAttempts, bucketOf, summarizeFixDurations, _internals, summarizeFixAxes, renderFixAxes */
 /* harmony import */ var node_fs__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(3024);
 /* harmony import */ var node_path__WEBPACK_IMPORTED_MODULE_1__ = __webpack_require__(6760);
 /* harmony import */ var _state_dir_js__WEBPACK_IMPORTED_MODULE_2__ = __webpack_require__(1174);
@@ -213,6 +213,82 @@ function renderFixDurationSummary(sum) {
 const _internals = { _dist, _pct, RELIABLE_N };
 
 
+// ── PRD F6.1 — score fixes on THREE AXES, not one ──────────────────────────
+//
+// The three axes the PRD names:
+//   (a) does the finding disappear      — the rescan leg
+//   (b) does the project's own suite pass — the tests leg
+//   (c) does an independent verifier agree — the PoC re-check leg
+//
+// All three were already computed by verifyFixCore and then collapsed into one
+// boolean, which is the problem: **(a) alone is satisfiable by deleting code.**
+// A patch that removes the vulnerable function passes the rescan, has nothing
+// left to fail, and — on a project with no detectable test suite — reaches
+// ok:true having proven only that the detector went quiet.
+//
+// Reporting the axes separately makes that visible. `aOnly` is the number that
+// matters most and the one nobody was publishing: attempts that satisfied ONLY
+// the disappearance axis. A high aOnly with a high headline is the shape of a
+// remediation feature that is deleting code and calling it a fix.
+function summarizeFixAxes(attempts) {
+  const list = Array.isArray(attempts) ? attempts.filter(Boolean) : [];
+  const d = list.length;
+
+  const rate = (pred) => ({ n: list.filter(pred).length, d });
+
+  // Each axis is judged INDEPENDENTLY of the overall verdict, so a leg that
+  // passed inside a failed attempt still counts for its own axis. Reading them
+  // off `ok` would make the three axes three copies of the same number.
+  const findingDisappeared = rate((a) => a.rescanOk === true || (a.ok === true && a.rescanOk !== false));
+  const testsStillPass = rate((a) => a.testsRan === true && a.testsOk !== false);
+  const verifierAgrees = rate((a) => a.pocOk === true);
+
+  const satisfiesAll = rate((a) =>
+    (a.rescanOk === true || (a.ok === true && a.rescanOk !== false))
+    && a.testsRan === true && a.testsOk !== false
+    && a.pocOk === true);
+
+  // The honesty number: disappearance WITHOUT either corroborating axis.
+  const aOnly = rate((a) => {
+    const disappeared = a.rescanOk === true || (a.ok === true && a.rescanOk !== false);
+    const corroborated = (a.testsRan === true && a.testsOk !== false) || a.pocOk === true;
+    return disappeared && !corroborated;
+  });
+
+  return {
+    total: d,
+    findingDisappeared,
+    testsStillPass,
+    verifierAgrees,
+    satisfiesAll,
+    aOnly,
+    meaning:
+      'findingDisappeared = the detector went quiet; testsStillPass = the project suite ran AND passed; '
+      + 'verifierAgrees = an independent PoC re-check confirmed the hole is shut. '
+      + 'aOnly counts attempts that satisfied ONLY disappearance — the shape a code-deleting "fix" produces.',
+    caveat: d === 0
+      ? 'no attempts recorded; every rate is 0/0 and means nothing'
+      : 'rates carry {n,d}; a small d is indicative, not settled',
+  };
+}
+
+/** Markdown for a report. Denominators always attached. */
+function renderFixAxes(sum) {
+  if (!sum || !sum.total) return '_No fix attempts recorded._\n';
+  const row = (label, r, note) => `| ${label} | ${r.n}/${r.d} | ${note} |`;
+  return [
+    '| Axis | Rate | Meaning |',
+    '|---|---|---|',
+    row('(a) finding disappeared', sum.findingDisappeared, 'the detector went quiet'),
+    row('(b) project tests pass', sum.testsStillPass, 'the suite RAN and passed'),
+    row('(c) verifier agrees', sum.verifierAgrees, 'an independent PoC re-check confirmed it'),
+    row('all three', sum.satisfiesAll, 'the only row that means "fixed"'),
+    row('(a) ALONE', sum.aOnly, 'satisfiable by deleting code — watch this number'),
+    '',
+  ].join('\n');
+}
+
+
 /***/ })
 
 };

package/dist/384.index.js

@@ -8,7 +8,7 @@ export const modules = {
 /* harmony export */ __webpack_require__.d(__webpack_exports__, {
 /* harmony export */   scanCredentials: () => (/* reexport safe */ _engine_js__WEBPACK_IMPORTED_MODULE_0__.Sv)
 /* harmony export */ });
-/* harmony import */ var _engine_js__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(6622);
+/* harmony import */ var _engine_js__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(3474);
 // Secrets submodule view of the engine — credential + entropy + TODO scanning.
 
 

package/dist/435.index.js

@@ -806,6 +806,18 @@ function _maybeOffload(sessionRoot, toolName, items) {
 }
 
 // ─── scan_diff ───────────────────────────────────────────────────────────────
+// Test seam for the write boundary (PRD F6.4).
+//
+// `_confine` and `isReservedWrite` ARE the confinement contract in
+// agents/_CONFINEMENT.md. A boundary is only worth what its refusals are worth,
+// and refusals cannot be adversarially tested through the public tools without
+// also exercising a real scan, a real patch and a real filesystem write — so
+// the check would be measuring four things and attributing failure to one.
+//
+// Exported under the `_internals` convention this codebase already uses
+// (see posture/poc-inprocess.js). Not part of the MCP tool surface.
+const tools_internals = { _confine, isReservedWrite: _isReservedWritePath };
+
 const scan_diff = {
   name: 'scan_diff',
   description: 'Scan a list of files for security findings. Use BEFORE writing a Write/Edit to disk so the agent can self-correct. Returns findings with severity, file:line, title, remediation. Snippets are redacted of obvious secret patterns. Paths confined to the session root; symlinks are refused.',