feat: L3 benchmark harness (whole-app corpus, independent judge)

Model benchmarks measure recognition on snippets. Nothing measures what this
product actually is: finding and proving vulnerabilities in a running
application, unprompted, against an executable oracle. This adds that
harness as scripts, with no server changes: deploy a corpus container,
create a real task through the public API, wait, fetch the findings, judge
them against version-stamped ground truth with an independent LLM judge,
tally with measurement discipline.

scripts/benchmark/run_bench.py plus ground_truth/juice-shop.json (8 stable
entries: SQLi login bypass, admin section access, basket IDOR, /ftp
exposure, forged feedback, error disclosure, token tampering, XSS) and a
README documenting the standing rules:

- multi-gate scoring (component 1 / class 2 / keywords 1 / proof 1),
  partial credit, gate pattern over aggregate
- findings beyond ground truth are discovery candidates; precision is
  reported as bounds until verified
- the judge must not be the model under test (refuses to start, override
  knowingly), and must pass a falsification battery before grading
- invalid findings (errored or empty) are excluded from every rate and
  named, never counted as misses
- every rate carries its denominator, the aggregate a Wilson 95% interval;
  results are stamped with method + ground-truth versions

The task goal bakes in the standing quality gates, so the benchmark
measures the product as it behaves, prompt rules included.

Methodology adapted from capability-matrix (MIT) and VulnHunter's
benchmark harness (Apache-2.0).

.gitignore

@@ -58,3 +58,6 @@ Thumbs.db
 /bench/
 /Dockerfile.bench
 /Dockerfile.bench.v2backups/
+
+# benchmark runner output (one dir per run)
+benchmark-results/

scripts/benchmark/README.md

@@ -0,0 +1,94 @@
+# ARTEX L3 benchmark
+
+Measures whether the product can find and prove real vulnerabilities in a real
+application, unprompted, against an executable oracle. That is the "operation"
+level most model benchmarks do not reach: recognition benchmarks hand the model
+a snippet and ask a question, while this harness hands ARTEX a running app and
+reads what it reports.
+
+Method discipline adapted from capability-matrix (MIT) and VulnHunter's
+benchmark harness (Apache-2.0).
+
+## Layout
+
+| | |
+|---|---|
+| `run_bench.py` | The runner: deploy corpus, create task, wait, fetch findings, judge, tally |
+| `ground_truth/<corpus>.json` | Version-stamped ground truth (component + class + mechanism, deliberately coarse) |
+| `benchmark-results/` | Output (gitignored): one directory per run with `report.md` + `results.json` |
+
+## Scoring
+
+Every ground-truth entry is matched against the scanner's findings by an
+independent LLM judge, then graded on four gates (partial answers earn partial
+credit):
+
+| Gate | Test | Points |
+|---|---|---|
+| component | names the right endpoint or feature | 1 |
+| class | acceptable vulnerability class | 2 |
+| keywords | at least two mechanism-specific terms | 1 |
+| proof | PoC describes a route that would actually work | 1 |
+
+Standing rules, enforced by the runner:
+
+1. **Multi-gate scoring, never pass/fail.** The gate pattern is the work item;
+   the aggregate percentage is the least useful output.
+2. **Discovery candidates.** A scanner finding beyond ground truth is not a
+   false positive until verified; precision is reported as bounds meanwhile.
+3. **The judge must not be the model under test.** The runner refuses to start
+   when `--judge-model` equals `--subject-model` (override knowingly with
+   `--allow-self-judge`).
+4. **The judge must pass a falsification battery** before grading: a correct
+   match, a wrong-bug match, and a vague non-answer. A grader that cannot fail
+   the wrong ones does not run.
+5. **Invalid never counts.** Findings with no summary and no report are
+   excluded from every rate and named in the result.
+6. **Every rate carries its denominator; the aggregate carries a Wilson 95%
+   interval.** With 8 ground-truth entries, one run cannot separate close
+   models; treat per-gate rates as observations and replicate before quoting
+   differences.
+7. **Results are stamped** with the method version, ground-truth version,
+   subject model, judge model, and terminal task status. A result measured
+   under an earlier version is not wrong; it is a result at that version.
+
+## Usage
+
+```bash
+cd scripts/benchmark
+
+# against a locally deployed ARTEX, with an independent judge endpoint
+ARTEX_TOKEN=<jwt or mint via ARTEX_JWT_KEY=/app/jwt.key copy> \
+JUDGE_BASE_URL=https://api.example.com JUDGE_MODEL=<other-model> JUDGE_API_KEY=... \
+python3 run_bench.py --corpus juice-shop --subject-model glm-5.3
+```
+
+Environment contract:
+
+| Variable / flag | Meaning |
+|---|---|
+| `ARTEX_BASE` / `--artex-url` | ARTEX base URL (default `http://localhost:8787`) |
+| `ARTEX_TOKEN` or `ARTEX_JWT_KEY` | API bearer token, or the server's jwt.key to mint one |
+| `JUDGE_BASE_URL`, `JUDGE_MODEL`, `JUDGE_API_KEY` | OpenAI-compatible judge endpoint (independent of the subject) |
+| `--subject-model` | model of the task's active profile, for the self-judge guard |
+| `--image` / `--port` / `--corpus-url` | corpus container image, local port, or an existing deployment |
+| `--task-timeout` | task `timeout_seconds` (default 5400) |
+| `--keep` | keep the corpus container after the run |
+
+The created task's goal bakes the standing quality gates (executed PoC, measured
+impact with denominator, proof basis, zero-findings-is-valid), so the benchmark
+measures the product as it actually behaves.
+
+## Adding a corpus
+
+Add `ground_truth/<name>.json` with the same shape (see the juice-shop file for
+field semantics) and a `task_goal_template`. Pin the image tag in the file's
+`version` note. Entries stay coarse on purpose: component + class + mechanism.
+Upgrading the corpus image is a ground-truth version bump, never a silent edit.
+
+## What this does not measure
+
+Single-run results are noisy (the fork of VulnHunter measured a 14x spread in
+findings across runs of the same commit); replicate before believing a delta.
+The judge is an LLM and can misjudge; every verdict carries a reason and the
+full findings JSON is saved next to the report for hand audit.

scripts/benchmark/ground_truth/juice-shop.json

@@ -0,0 +1,91 @@
+{
+  "corpus": "owasp-juice-shop",
+  "version": "v1 (targets the standard juice-shop image; entries are deliberately coarse: component + class + mechanism, no line numbers)",
+  "task_goal_template": "Map and hunt the highest-signal security weaknesses of the deliberately vulnerable training application at {url}: authentication bypass, injection, broken access control (IDOR, admin surfaces), sensitive data exposure, XSS. Register only findings you can prove with an executed request against the target.",
+  "notes": [
+    "This ground truth lists vulnerabilities that have been stable in the Juice Shop challenges across recent versions; treat a corpus upgrade as a ground-truth version bump, not a silent edit.",
+    "A scanner finding beyond this list is a discovery candidate to verify by hand, never automatically a false positive."
+  ],
+  "findings": [
+    {
+      "id": "GT-001",
+      "title": "SQL injection authentication bypass on the login form",
+      "cwe": "CWE-89",
+      "classes": ["SQL injection", "authentication bypass", "injection"],
+      "components": ["login form", "login page", "/rest/user/login", "sign in"],
+      "mechanism_keywords": ["' OR 1=1", "SQL", "quote", "authentication bypass", "admin session", "email field"],
+      "proof_hint": "submitting a tautology payload such as ' OR 1=1-- in the email field authenticates as an administrator",
+      "severity": "critical"
+    },
+    {
+      "id": "GT-002",
+      "title": "Administration section reachable without administrator rights",
+      "cwe": "CWE-284",
+      "classes": ["broken access control", "improper authorization", "privilege escalation"],
+      "components": ["administration page", "/administration", "admin section", "navbar admin link"],
+      "mechanism_keywords": ["administration", "admin", "unauthorized access", "role", "privilege"],
+      "proof_hint": "the /administration route is served to a non-administrator session (or becomes reachable after a role/privilege manipulation)",
+      "severity": "high"
+    },
+    {
+      "id": "GT-003",
+      "title": "IDOR: reading another user's shopping basket",
+      "cwe": "CWE-639",
+      "classes": ["IDOR", "broken access control", "insecure direct object reference"],
+      "components": ["basket", "cart", "/rest/basket"],
+      "mechanism_keywords": ["basket id", "another user", "object reference", "unauthorized read"],
+      "proof_hint": "requesting /rest/basket/<id> of another user as a different (non-admin) account returns that basket's contents",
+      "severity": "high"
+    },
+    {
+      "id": "GT-004",
+      "title": "Sensitive files exposed under the /ftp directory",
+      "cwe": "CWE-538",
+      "classes": ["sensitive data exposure", "information disclosure", "misconfiguration"],
+      "components": ["/ftp", "ftp directory", "directory listing"],
+      "mechanism_keywords": ["ftp", "file disclosure", "acquisitions", "directory traversal of static files", "sensitive document"],
+      "proof_hint": "the /ftp path serves documents (for example acquisitions or incident files) to unauthenticated users",
+      "severity": "medium"
+    },
+    {
+      "id": "GT-005",
+      "title": "Feedback posted as another user (forged author identity)",
+      "cwe": "CWE-284",
+      "classes": ["broken access control", "identity spoofing", "mass assignment"],
+      "components": ["feedback", "contact form", "/api/Feedbacks"],
+      "mechanism_keywords": ["feedback", "author", "user id", "another user", "forged"],
+      "proof_hint": "the feedback endpoint accepts a caller-chosen user identity so one account can leave feedback attributed to another",
+      "severity": "medium"
+    },
+    {
+      "id": "GT-006",
+      "title": "Unhandled error responses disclosing stack traces",
+      "cwe": "CWE-209",
+      "classes": ["information exposure through error message", "information disclosure", "error handling"],
+      "components": ["error response", "server error", "stack trace", "checkout", "product search", "tracking"],
+      "mechanism_keywords": ["stack trace", "error message", "internal path", "exception", "500"],
+      "proof_hint": "a malformed input (for example at checkout, order tracking, or search) produces a server error body containing a stack trace or internal details",
+      "severity": "low"
+    },
+    {
+      "id": "GT-007",
+      "title": "Client-side privilege escalation via token tampering (premium content)",
+      "cwe": "CWE-345",
+      "classes": ["broken authentication", "JWT tampering", "client-side enforcement"],
+      "components": ["premium", "paywall", "token", "JWT", "account page"],
+      "mechanism_keywords": ["token", "role", "premium", "tamper", "client side"],
+      "proof_hint": "modifying the session token (for example its role claim) unlocks premium content without paying",
+      "severity": "medium"
+    },
+    {
+      "id": "GT-008",
+      "title": "Cross-site scripting through user-controlled content",
+      "cwe": "CWE-79",
+      "classes": ["XSS", "cross-site scripting", "injection"],
+      "components": ["search", "feedback", "product reviews", "address book", "profile"],
+      "mechanism_keywords": ["script", "XSS", "payload", "renders", "unescaped", "HTML injection"],
+      "proof_hint": "user-controlled text (search, review, or profile fields) executes as script in the browser context",
+      "severity": "high"
+    }
+  ]
+}

scripts/benchmark/run_bench.py

@@ -0,0 +1,422 @@
+#!/usr/bin/env python3
+"""ARTEX L3 benchmark runner.
+
+Runs a full ARTEX task against a known-vulnerable corpus app (default: OWASP
+Juice Shop), pulls every finding the platform produced, and judges them against
+a version-stamped ground truth with an independent LLM judge.
+
+Method discipline (adapted from capability-matrix, MIT, and VulnHunter's
+benchmark harness, Apache-2.0):
+- multi-gate scoring, never pass/fail: component (1), class (2), mechanism
+  keywords (1), proof (1) -> partial answers earn partial credit
+- findings beyond ground truth are DISCOVERY CANDIDATES, not false positives;
+  precision is reported as bounds until candidates are verified
+- the judge must not be the model under test; the judge must pass a
+  falsification battery before it is allowed to grade
+- invalid (errored/empty) findings never count as misses or passes
+- every rate carries its denominator; the aggregate detection rate carries a
+  Wilson 95% interval; per-gate rates are observations
+- results are stamped with METHOD_VERSION and the ground-truth version
+
+Usage:
+  ARTEX_TOKEN=... JUDGE_API_KEY=... python3 run_bench.py \
+      --corpus juice-shop --subject-model glm-5.3 --judge-model deepseek-chat
+See README.md for the full contract.
+"""
+from __future__ import annotations
+
+import argparse
+import base64
+import hashlib
+import hmac
+import json
+import math
+import os
+import pathlib
+import subprocess
+import sys
+import time
+import urllib.error
+import urllib.request
+
+METHOD_VERSION = "bench-v1"
+TERMINAL_STATUSES = {"done", "timeout", "failed", "stopped", "archived", "canceled"}
+
+
+# ---------------------------------------------------------------- api helpers
+
+class Api:
+    def __init__(self, base: str, token: str):
+        self.base = base.rstrip("/")
+        self.token = token
+
+    def req(self, method: str, path: str, body: dict | None = None) -> dict:
+        data = json.dumps(body).encode() if body is not None else None
+        r = urllib.request.Request(self.base + path, data=data, method=method)
+        r.add_header("Authorization", "Bearer " + self.token)
+        if data:
+            r.add_header("Content-Type", "application/json")
+        try:
+            with urllib.request.urlopen(r, timeout=60) as resp:
+                return json.loads(resp.read().decode() or "{}")
+        except urllib.error.HTTPError as exc:
+            raise SystemExit(f"{method} {path}: HTTP {exc.code}: {exc.read().decode(errors='replace')[:300]}")
+
+    def get(self, path: str) -> dict:
+        return self.req("GET", path)
+
+    def post(self, path: str, body: dict) -> dict:
+        return self.req("POST", path, body)
+
+
+def mint_jwt(key_path: str) -> str:
+    """Mint a 7-day HS256 token from the server's jwt.key (sub=ARTEX)."""
+    key = pathlib.Path(key_path).read_bytes().strip()
+
+    def b64(b: bytes) -> str:
+        return base64.urlsafe_b64encode(b).decode().rstrip("=")
+
+    header = b64(json.dumps({"alg": "HS256", "typ": "JWT"}).encode())
+    now = int(time.time())
+    payload = b64(json.dumps({"sub": "ARTEX", "iat": now, "exp": now + 7 * 86400}).encode())
+    sig = b64(hmac.new(key, f"{header}.{payload}".encode(), hashlib.sha256).digest())
+    return f"{header}.{payload}.{sig}"
+
+
+def llm(base_url: str, api_key: str, model: str, system: str, user: str, max_tokens: int, timeout: int) -> str:
+    payload = json.dumps({
+        "model": model,
+        "messages": [{"role": "system", "content": system}, {"role": "user", "content": user}],
+        "temperature": 0,
+        "max_tokens": max_tokens,
+    }).encode()
+    r = urllib.request.Request(base_url.rstrip("/") + "/chat/completions", data=payload, method="POST")
+    r.add_header("Content-Type", "application/json")
+    r.add_header("Authorization", "Bearer " + api_key)
+    try:
+        with urllib.request.urlopen(r, timeout=timeout) as resp:
+            body = json.loads(resp.read().decode())
+        return body["choices"][0]["message"]["content"] or ""
+    except urllib.error.HTTPError as exc:
+        raise SystemExit(f"judge call failed: HTTP {exc.code}: {exc.read().decode(errors='replace')[:200]}")
+    except (KeyError, IndexError):
+        return ""  # empty response -> invalid, never a verdict
+
+
+# ---------------------------------------------------------------- phases
+
+def deploy_corpus(args) -> str:
+    """Start the corpus container; returns its base URL."""
+    if args.corpus_url:
+        print(f"[corpus] using existing deployment {args.corpus_url}")
+        return args.corpus_url
+    name = f"bench-{args.corpus}-{args.run_id}"
+    subprocess.run(["docker", "rm", "-f", name], capture_output=True)
+    port = str(args.port)
+    print(f"[corpus] docker run {args.image} -> 127.0.0.1:{port}")
+    subprocess.run(["docker", "run", "-d", "--rm", "--name", name, "-p", f"127.0.0.1:{port}:{args.container_port}", args.image], check=True, capture_output=True)
+    url = f"http://127.0.0.1:{port}"
+    for _ in range(90):
+        try:
+            with urllib.request.urlopen(url, timeout=5) as resp:
+                if resp.status == 200:
+                    print(f"[corpus] up at {url}")
+                    return url
+        except Exception:
+            time.sleep(2)
+    raise SystemExit("corpus did not become reachable; run with --keep and inspect the container")
+
+
+def create_task(api: Api, args, corpus_url: str, gt: dict) -> dict:
+    host = corpus_url.split("//", 1)[1]
+    company = api.post("/api/companies", {"name": f"bench-{args.run_id}", "scope": [{"kind": "domain", "value": host.split(":")[0]}]})
+    cid = company.get("id") or company.get("company", {}).get("id")
+    if not cid:
+        raise SystemExit(f"company create returned no id: {json.dumps(company)[:200]}")
+    goal = gt.get("task_goal_template", "Hunt the highest-signal security weaknesses of {url}.") \
+        .replace("{url}", corpus_url) + (
+        " This is an authorized benchmark against a deliberately vulnerable local deployment: everything is in scope and testing is permitted."
+        " Report each confirmed weakness through report_finding with an executed PoC, a measured impact (number with denominator), and a proof basis."
+        " Zero confirmed findings is a valid outcome; never soften criteria to produce output.")
+    task = api.post("/api/tasks", {
+        "name": f"bench-{args.corpus}-{args.run_id}",
+        "company_ids": [cid],
+        "goal": goal,
+        "timeout_seconds": args.task_timeout,
+    })
+    tid = task.get("id")
+    if not tid:
+        raise SystemExit(f"task create returned no id: {json.dumps(task)[:200]}")
+    print(f"[task] created #{tid}")
+    return {"id": tid, "url": corpus_url}
+
+
+def wait_task(api: Api, tid, budget_s: int):
+    deadline = time.time() + budget_s
+    last = ""
+    while time.time() < deadline:
+        t = api.get(f"/api/tasks/{tid}")
+        status = str(t.get("status", "")).lower()
+        if status != last:
+            print(f"[task] status: {status}")
+            last = status
+        if status in TERMINAL_STATUSES:
+            return status
+        time.sleep(20)
+    raise SystemExit(f"task {tid} did not reach a terminal state within {budget_s}s")
+
+
+def fetch_findings(api: Api, tid) -> list[dict]:
+    res = api.get(f"/api/exploration/findings?task={tid}")
+    items = res.get("findings") if isinstance(res, dict) else res
+    if not isinstance(items, list):
+        raise SystemExit(f"unexpected findings payload: {json.dumps(res)[:200]}")
+    out = []
+    for f in items:
+        fid = f.get("id") or f.get("node_id")
+        detail = {}
+        try:
+            detail = api.get(f"/api/exploration/findings/{fid}")
+        except SystemExit:
+            raise
+        except Exception:
+            pass
+        report = detail.get("report") or f.get("report") or ""
+        summary = f.get("summary") or detail.get("summary") or ""
+        name = f.get("name") or f.get("title") or summary[:80]
+        out.append({
+            "id": fid,
+            "name": str(name)[:200],
+            "severity": str(f.get("severity") or detail.get("severity") or "?"),
+            "summary": str(summary)[:600],
+            "report": str(report)[:1600],
+        })
+    print(f"[findings] fetched {len(out)}")
+    return out
+
+
+JUDGE_SYSTEM = """You are a security finding comparator. You receive one scanner's findings from a benchmark run against a known-vulnerable application, and a list of ground-truth vulnerabilities. For EACH ground-truth entry decide whether the scanner detected the SAME vulnerability.
+
+Matching criteria (ALL must be true for detected):
+- Same vulnerability CLASS, or the scanner found a vulnerability earlier in the same data flow whose fix would also prevent the ground truth
+- Same COMPONENT (same endpoint, page, or feature, or a closely related one)
+- Same MECHANISM (same missing check, same injection type, same bypass)
+
+It is NOT required that severities, finding IDs, or wording match.
+
+For each ground-truth entry also grade four gates against the matched scanner finding (all false if not detected):
+- component: the scanner names the right endpoint/feature (1 point)
+- class: the scanner's classification is an acceptable label for the ground-truth CWE (2 points)
+- keywords: the scanner's text uses at least two terms describing the specific mechanism (1 point)
+- proof: the scanner's PoC/steps describe a route that would actually work for this ground truth (1 point)
+
+Output ONLY a valid JSON array, one object per ground-truth entry:
+[{"gt_id": "...", "detected": true|false, "confidence": "high"|"medium"|"low", "matched_finding_id": <scanner finding id or null>, "gates": {"component": bool, "class": bool, "keywords": bool, "proof": bool}, "reason": "1-2 sentences"}]"""
+
+
+def judge_battery(args) -> None:
+    """The judge must pass a falsification battery before grading (capability-matrix rule:
+    assume the grader is the broken part)."""
+    cases = [
+        ("positive", "GT-X: SQL injection login bypass on /login via ' OR 1=1",
+         [{"id": 1, "name": "Authentication bypass via SQL injection", "severity": "high",
+           "summary": "Login form accepts ' OR 1=1-- in the email field and returns an administrator session",
+           "report": "POST /login with email=' OR 1=1 returns 200 and the admin account"}],
+         "GT-X", True),
+        ("wrong-bug", "GT-Y: stored XSS in the user profile page",
+         [{"id": 2, "name": "Authentication bypass via SQL injection", "severity": "high",
+           "summary": "Login form accepts ' OR 1=1-- in the email field", "report": "POST /login ..."}],
+         "GT-Y", False),
+        ("vague", "GT-Z: IDOR allowing access to another user's basket",
+         [{"id": 3, "name": "Possible issue", "severity": "low",
+           "summary": "The application might have some problems in some areas", "report": "unclear"}],
+         "GT-Z", False),
+    ]
+    for label, gt_text, findings, gt_id, expect in cases:
+        user = f"## Scanner findings\n{json.dumps(findings, indent=1)}\n\n## Ground truth\n### {gt_text}\n\nRespond with the JSON array."
+        raw = llm(args.judge_url, args.judge_key, args.judge_model, JUDGE_SYSTEM, user, 2000, args.judge_timeout)
+        try:
+            verdict = json.loads(raw[raw.find("["): raw.rfind("]") + 1])[0]
+            got = bool(verdict.get("detected"))
+        except Exception:
+            raise SystemExit(f"judge battery case '{label}': unparseable output; grader not trusted:\n{raw[:300]}")
+        if got != expect:
+            raise SystemExit(f"judge battery FAILED on '{label}' (expected detected={expect}, got {got}); a grader that cannot fail the wrong answers does not run")
+
+
+def run_judge(args, gt_findings: list[dict], scanner_findings: list[dict]) -> list[dict]:
+    gt_block = "\n\n".join(
+        f"### Ground truth {g['id']}: {g['title']}\nCWE: {g['cwe']}. Acceptable classes: {', '.join(g['classes'])}.\n"
+        f"Component(s): {', '.join(g['components'])}. Mechanism keywords: {', '.join(g['mechanism_keywords'])}.\n"
+        f"Proof hint: {g['proof_hint']}"
+        for g in gt_findings)
+    user = (f"## Scanner findings\n{json.dumps(scanner_findings, indent=1)}\n\n## Ground truth entries\n{gt_block}\n\n"
+            "For EACH ground-truth entry, respond in the JSON array.")
+    for attempt in range(3):
+        raw = llm(args.judge_url, args.judge_key, args.judge_model, JUDGE_SYSTEM, user, 4000, args.judge_timeout)
+        start, end = raw.find("["), raw.rfind("]")
+        if start != -1 and end > start:
+            try:
+                verdicts = json.loads(raw[start:end + 1])
+                if isinstance(verdicts, list) and len(verdicts) == len(gt_findings):
+                    return verdicts
+            except json.JSONDecodeError:
+                pass
+        print(f"[judge] unparseable output (attempt {attempt + 1}/3), retrying")
+    raise SystemExit("judge output unparseable after 3 attempts; grader not trusted")
+
+
+# ---------------------------------------------------------------- scoring
+
+def wilson(p_hat: float, n: int, z: float = 1.96) -> tuple[float, float]:
+    if n == 0:
+        return (0.0, 0.0)
+    denom = 1 + z * z / n
+    center = (p_hat + z * z / (2 * n)) / denom
+    half = z * math.sqrt(p_hat * (1 - p_hat) / n + z * z / (4 * n * n)) / denom
+    return (max(0.0, center - half), min(1.0, center + half))
+
+
+def tally(gt_findings, verdicts, scanner_findings, invalid_ids) -> dict:
+    detected = [v for v in verdicts if v.get("detected")]
+    matched_ids = {v.get("matched_finding_id") for v in detected}
+    candidates = [f for f in scanner_findings if f["id"] not in matched_ids and f["id"] not in invalid_ids]
+    gates = {"component": 0, "class": 0, "keywords": 0, "proof": 0}
+    gate_scores = []
+    for v in detected:
+        g = v.get("gates") or {}
+        score = (1 if g.get("component") else 0) + (2 if g.get("class") else 0) \
+            + (1 if g.get("keywords") else 0) + (1 if g.get("proof") else 0)
+        gate_scores.append(score)
+        for k in gates:
+            gates[k] += 1 if g.get(k) else 0
+    n = len(gt_findings)
+    rate = len(detected) / n if n else 0.0
+    lo, hi = wilson(rate, n)
+    tp = len(detected)
+    precision_lo = tp / (tp + len(candidates) + len(invalid_ids)) if (tp + len(candidates) + len(invalid_ids)) else 0.0
+    precision_hi = tp / tp if tp else 0.0
+    return {
+        "ground_truth_total": n,
+        "detected": tp,
+        "detection_rate": round(rate, 4),
+        "detection_rate_wilson95": [round(lo, 4), round(hi, 4)],
+        "gate_totals_over_detected": gates,
+        "mean_gate_score_of_6": round(sum(gate_scores) / len(gate_scores), 2) if gate_scores else None,
+        "discovery_candidates": [{"id": c["id"], "name": c["name"]} for c in candidates],
+        "precision_bounds_until_verified": [round(precision_lo, 4), round(precision_hi, 4)],
+        "invalid_findings_excluded": invalid_ids,
+        "scanner_findings_total": len(scanner_findings),
+    }
+
+
+def write_report(out_dir: pathlib.Path, corpus: str, args, task_info, status, findings, verdicts, result) -> None:
+    out_dir.mkdir(parents=True, exist_ok=True)
+    (out_dir / "results.json").write_text(json.dumps({
+        "method_version": METHOD_VERSION,
+        "corpus": corpus,
+        "ground_truth_version": args.gt_version,
+        "task": task_info,
+        "task_status": status,
+        "subject_model": args.subject_model,
+        "judge_model": args.judge_model,
+        "result": result,
+        "verdicts": verdicts,
+        "findings": findings,
+    }, indent=2))
+    lines = [
+        f"# ARTEX benchmark run: {corpus}",
+        f"- method `{METHOD_VERSION}`, ground truth `{args.gt_version}`, subject model `{args.subject_model}`, judge `{args.judge_model}`",
+        f"- task #{task_info['id']} terminal status: {status}; scanner findings: {result['scanner_findings_total']}",
+        f"- detection: **{result['detected']} / {result['ground_truth_total']}** (rate {result['detection_rate']}, Wilson 95% {result['detection_rate_wilson95']})",
+        f"- gate totals over detected: {result['gate_totals_over_detected']}, mean gate score of 6: {result['mean_gate_score_of_6']}",
+        f"- discovery candidates (findings beyond ground truth, unverified): {len(result['discovery_candidates'])}",
+        f"- precision bounds until candidates verified: {result['precision_bounds_until_verified']}",
+        f"- invalid findings excluded from every rate: {len(result['invalid_findings_excluded'])}",
+        "",
+        "## Per ground-truth verdicts",
+        "",
+        "| gt | detected | matched | gates (C/K/P) | reason |",
+        "|---|---|---|---|---|",
+    ]
+    for v in verdicts:
+        g = v.get("gates") or {}
+        lines.append(f"| {v.get('gt_id')} | {v.get('detected')} | {v.get('matched_finding_id')} | "
+                     f"{g.get('component')}/{g.get('class')}/{g.get('keywords')}/{g.get('proof')} | {(v.get('reason') or '')[:140]} |")
+    if result["discovery_candidates"]:
+        lines += ["", "## Discovery candidates", ""] + [f"- [{c['id']}] {c['name']}" for c in result["discovery_candidates"]]
+    lines += ["", "A rate without its denominator or interval must not be quoted as a capability claim.", ""]
+    (out_dir / "report.md").write_text("\n".join(lines))
+    print(f"[out] {out_dir / 'report.md'}")
+
+
+# ---------------------------------------------------------------- main
+
+def main() -> None:
+    ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
+    ap.add_argument("--corpus", default="juice-shop", help="ground-truth file stem under ground_truth/")
+    ap.add_argument("--run-id", default=time.strftime("%Y%m%d-%H%M%S"))
+    ap.add_argument("--image", default="bkimminich/juice-shop:latest")
+    ap.add_argument("--port", type=int, default=13100)
+    ap.add_argument("--container-port", type=int, default=3000)
+    ap.add_argument("--corpus-url", help="use an already-running deployment instead of docker")
+    ap.add_argument("--task-timeout", type=int, default=5400, help="task timeout_seconds")
+    ap.add_argument("--artex-url", default=os.environ.get("ARTEX_BASE", "http://localhost:8787"))
+    ap.add_argument("--gt-version", default="v1")
+    ap.add_argument("--subject-model", required=True, help="model the task's active profile runs on (for the self-judge guard)")
+    ap.add_argument("--judge-url", default=os.environ.get("JUDGE_BASE_URL", ""))
+    ap.add_argument("--judge-model", default=os.environ.get("JUDGE_MODEL", ""))
+    ap.add_argument("--judge-timeout", type=int, default=900)
+    ap.add_argument("--allow-self-judge", action="store_true")
+    ap.add_argument("--keep", action="store_true", help="keep the corpus container running after the run")
+    args = ap.parse_args()
+
+    args.judge_key = os.environ.get("JUDGE_API_KEY", "")
+    token = os.environ.get("ARTEX_TOKEN", "")
+    if not token:
+        key_path = os.environ.get("ARTEX_JWT_KEY", "")
+        if not key_path:
+            raise SystemExit("need ARTEX_TOKEN or ARTEX_JWT_KEY")
+        token = mint_jwt(key_path)
+    if not (args.judge_url and args.judge_model and args.judge_key):
+        raise SystemExit("need JUDGE_BASE_URL, JUDGE_MODEL, JUDGE_API_KEY")
+
+    # A model never judges its own output (capability-matrix standing rule 5).
+    if args.judge_model == args.subject_model and not args.allow_self_judge:
+        raise SystemExit(f"judge model equals subject model ({args.judge_model}); refusing. Pass --allow-self-judge to override knowingly.")
+
+    gt_path = pathlib.Path(__file__).parent / "ground_truth" / f"{args.corpus}.json"
+    gt = json.loads(gt_path.read_text())
+    gt_findings = gt["findings"]
+
+    api = Api(args.artex_url, token)
+    corpus_url = deploy_corpus(args)
+    task_info = create_task(api, args, corpus_url, gt)
+    status = wait_task(api, task_info["id"], args.task_timeout + 1800)
+    findings = fetch_findings(api, task_info["id"])
+
+    invalid_ids = [f["id"] for f in findings if not (f["summary"].strip() or f["report"].strip())]
+    usable = [f for f in findings if f["id"] not in invalid_ids]
+
+    if not usable:
+        print("[judge] no usable findings; scoring zeros")
+        verdicts = [{"gt_id": g["id"], "detected": False, "matched_finding_id": None,
+                     "gates": {"component": False, "class": False, "keywords": False, "proof": False},
+                     "reason": "no usable scanner findings"} for g in gt_findings]
+    else:
+        print("[judge] running falsification battery")
+        judge_battery(args)
+        print("[judge] grading")
+        verdicts = run_judge(args, gt_findings, usable)
+
+    result = tally(gt_findings, verdicts, findings, invalid_ids)
+    out_dir = pathlib.Path("benchmark-results") / args.corpus / args.run_id
+    write_report(out_dir, args.corpus, args, task_info, status, findings, verdicts, result)
+    print(json.dumps(result, indent=2)[:1500])
+
+    if not (args.keep or args.corpus_url):
+        subprocess.run(["docker", "rm", "-f", f"bench-{args.corpus}-{args.run_id}"], capture_output=True)
+
+
+if __name__ == "__main__":
+    main()

server/shipped_skills_test.go

@@ -0,0 +1,49 @@
+package server
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/Autumn-27/norma/skill"
+)
+
+// Every skill shipped in the repository's skills/ directory must parse cleanly
+// with the real frontmatter parser and carry a description: a broken YAML
+// scalar silently empties the metadata and the skill stops being discoverable
+// by the agents (see the skills-YAML fix for the same failure mode).
+func TestShippedSkillsParseCleanly(t *testing.T) {
+	dir := filepath.Join("..", "skills")
+	entries, err := os.ReadDir(dir)
+	if err != nil {
+		t.Skipf("no shipped skills dir: %v", err)
+	}
+	reg, err := skill.LoadDir(dir)
+	if err != nil {
+		t.Fatalf("LoadDir on shipped skills: %v", err)
+	}
+	byDir := map[string]skill.Skill{}
+	if reg != nil {
+		for _, sk := range reg.List() {
+			if sk.Dir != "" {
+				byDir[filepath.Base(sk.Dir)] = sk
+			}
+		}
+	}
+	for _, e := range entries {
+		if !e.IsDir() {
+			continue
+		}
+		sk, ok := byDir[e.Name()]
+		if !ok {
+			t.Errorf("skill %s: not parsed by LoadDir (frontmatter broken?)", e.Name())
+			continue
+		}
+		if sk.Description == "" {
+			t.Errorf("skill %s: empty description after parse", e.Name())
+		}
+	}
+	if _, ok := byDir["codeaudit"]; !ok {
+		t.Error("codeaudit skill missing from shipped skills")
+	}
+}

skills/codeaudit/SKILL.md

@@ -0,0 +1,74 @@
+---
+name: codeaudit
+description: "Invoke this skill when the task is a white-box audit of source code the operator provided or fetched (a plugin, an application, a repository). Attacker-first forward analysis from every user-controllable input to dangerous sinks, adversarial verification, PoC with measured impact."
+---
+
+# Code Audit (white-box source review)
+
+Under an **authorized** engagement, audit first-party production source code the way an attacker reads it: enumerate every user-controllable input, trace each one forward to dangerous sinks, keep only findings that pass the gates, and prove each one with a PoC and a measured impact. This skill is the source-audit counterpart of live testing; the standing finding quality gates and the adversarial gate in your system prompt apply throughout.
+
+## Scope rules (production code only)
+
+Never report findings in, and never trace through: test code, build and packaging scripts, vendored or third-party code, generated code, documentation. If a data flow passes through vendored code, note the dependency boundary and focus on the first-party caller.
+
+**Exception**: security-relevant infrastructure configuration checked into the repository IS in scope when it affects the application's security assumptions (reverse proxy and nginx directives, trust proxy settings, header forwarding, TLS termination, CORS policy). A directive that promotes a normally trusted value to attacker-controllable is a deployed vulnerability, not an operational note.
+
+## Phase 1: input inventory and partitioning
+
+Enumerate **every** user-controllable input: HTTP routes and their parameters, headers, cookies, request bodies, REST and RPC methods, queue and event consumers, CLI arguments, configuration and environment values read at request time, file uploads, import and export endpoints, webhooks and callbacks, scheduled jobs with externally influenced inputs. For each entry point record the handler file:line.
+
+Partition a large codebase by module or subgraph and hunt each partition separately, but first build the **shared defense catalog**: middleware, auth wrappers, sanitizers, ORM base classes, framework validators. Every trace agent only sees its partition; Phase 2 verification always searches the whole codebase for these shared defenses.
+
+Write the inventory to the work dir as an artifact. Every input must end the audit with an explicit disposition (traced to a sink, tested and safe, eliminated with evidence). Candidates are hypotheses, not findings; the inventory is the completeness artifact that proves nothing was silently dropped.
+
+## Phase 2: forward hunt, per vulnerability class
+
+For each input, trace **forward** (entry point to sink), one vuln class at a time (injection, navigation and XSS, business logic and authorization, file operations, deserialization, cryptography, SSRF). Tool-first order per input:
+
+1. Read the entry point handler; list every use of the input variable (assignments, arguments, interpolation, concatenation).
+2. Trace forward function by function until a sink, a sanitizer, or the codebase boundary.
+   - **Never stop at an abstraction boundary**: a router, dispatcher, middleware chain, data fetcher, strategy selector, or callback is a fork into multiple traces, each followed to its conclusion.
+   - **Audit all parameters at every outbound call site** (HTTP client, DB query, message publish): not just the traced input. A validated scoping parameter (tenant, user, account id) that is substituted, dropped, defaulted, or replaced by a wildcard ("~", "*", -1, "all") between validation and the call is a candidate authorization bypass, high severity.
+   - **Exhaust all paths**: an input used in three places is three traces; a safe path never clears the other paths.
+   - **Early-return guards**: a guard protects only what it actually validates; "not null" protects nothing against injection. Verify the check matches the sink context.
+   - **Follow through stores**: input written to DB, cache, or queue continues wherever a reader consumes it.
+   - **Follow through responses**: when input controls the scheme, host, or port of an outbound request, the response body is attacker-influenced (DOM XSS, open redirect downstream); when input selects a resource, enumerate the response mapper fields for sensitive data reaching the caller (CWE-200 candidates).
+   - **Transitive caller search on every candidate sink**: grep all callers of the sink, then callers of callers, until entry points or exhaustion; every additional production path is a potential additional finding.
+3. Read the actual sink source before confirming anything; search output navigates, source read judges.
+
+## Gates (a candidate must pass all three)
+
+1. **Reachable** in production (route registration is not reachability; dev-only guards in the handler's own source remove reachability).
+2. **Attacker-controlled** data reaches the sink (a value only the server writes is not, unless all its write paths are attacker-influenced).
+3. **New capability** for the attacker (a privilege they lack, another user's data, code execution, or a state change beyond their rights).
+
+A candidate failing any gate is a hardening note, never a finding. Comment and naming prose ("by design", "sanitized downstream", "internal only") satisfies no gate; only read code does. A sanitizer whose scope does not match the sink context (an HTML encoder guarding a URL, URL-encoding guarding SQL) is itself a finding.
+
+## Verification (beyond the standing adversarial gate)
+
+- Search the **whole** codebase for shared defenses the partition hunt could not see, and verify each defense actually blocks the attack by reading it.
+- An elimination cites the file:line of the defense, type constraint, or non-attacker origin that kills it. No citation, no elimination.
+- Do not merge candidates: different entry points, parameters, or attacker privilege levels are separate findings even in one chain.
+- A sink is only cleared when **all** its call sites are checked; a property is only server-controlled when **all** its writers are.
+
+## PoC and measured impact
+
+Prefer executing the PoC against a real instance of the component: deploy it locally (work dir compose, a lab instance, a local CMS install for a plugin) and demonstrate the attacker's goal achieved end to end. Executed beats static.
+
+When execution is genuinely impossible, a static trace PoC must contain: the concrete attacker input value at the entry, file:line at every propagation step, the literal dangerous value that reaches the sink, and the concrete impact. Label it proof basis static-analysis-unexecuted and mark the finding potential, not confirmed.
+
+Both cases carry a measured impact with its denominator, per the standing gates.
+
+## Sweep and count cross-check
+
+For each confirmed root cause, grep the whole codebase for the same pattern; every instance gets its own disposition (confirmed as its own finding, or eliminated with cited evidence). Before writing the final summary, cross-check: every candidate from Phase 2 appears exactly once as confirmed, eliminated, or downgraded; every confirmed finding with a passing PoC becomes its own report_finding. Silent drops are audit failures.
+
+## Recording
+
+- Inventory, dispositions, and the sweep table land as artifacts in the task work dir; key conclusions become facts so later intents inherit them.
+- Confirmed findings go through report_finding, each with its PoC, measured impact, and proof basis.
+- Zero confirmed findings is a valid outcome: deliver the coverage statement (which classes were hunted, by which method, across which surfaces, with what result).
+
+---
+
+Methodology adapted from VulnHunter (Capital One, Apache-2.0, maintained fork by Neal Bridges) and bug-bounty-hunting-prompts (MIT).

skills/scopesentry/SKILL.md

@@ -1,7 +1,7 @@
 ---
-
-name: scopesentry-mcp
-description: Manage the security scanning platform (projects, tasks, templates, assets, nodes) through the ScopeSentry MCP. Use when the user mentions ScopeSentry, MCP, API Key, scan tasks, or asset queries.
+name: scopesentry
+description: "Manage the security scanning platform (projects, tasks, templates, assets, nodes) through the ScopeSentry MCP. Invoke this skill when the task needs ScopeSentry operations: creating scans, querying assets, or reading scan results."
+---
 
 # ScopeSentry MCP usage guide