Publish Layerfault deep challenge LF-CH-TOKX-0014

CHALLENGE_ORACLE.json

@@ -0,0 +1,28 @@
+{
+  "attack_surface": [
+    "tokenizer-processor"
+  ],
+  "control_type": "negative",
+  "corpus_id": "LF-CH-TOKX-0014",
+  "difficulty": "compound",
+  "expected_decision": "PASS",
+  "ground_truth": "synthetic-generated",
+  "layerfault_rule_expectations": {
+    "candidate_rules": [],
+    "expected_rules": [],
+    "must_not_rules": []
+  },
+  "oracle_ids": [
+    "LF-ORACLE-TOKX-0014"
+  ],
+  "related_cases": [],
+  "repo_name": "tokenizer-security-words-control",
+  "safe_fixture": true,
+  "schema": 2,
+  "severity": "informational",
+  "techniques": [
+    "tokenizer",
+    "control"
+  ],
+  "transformations": []
+}

FIXTURE_LAYOUT.json

@@ -0,0 +1,9 @@
+{
+  "artifacts": [
+    "tokenizer_config.json",
+    "SECURITY.md"
+  ],
+  "family": "tokenizer",
+  "note": "All content is synthetic and inert. Do not treat this as a production model.",
+  "repo_name": "tokenizer-security-words-control"
+}

LAYERFAULT_CORPUS.json

@@ -0,0 +1,36 @@
+{
+  "candidate_rules": [],
+  "challenge": {
+    "attack_surface": [
+      "tokenizer-processor"
+    ],
+    "control_type": "negative",
+    "difficulty": "compound",
+    "expected_decision": "PASS",
+    "oracle_ids": [
+      "LF-ORACLE-TOKX-0014"
+    ],
+    "related_cases": [],
+    "severity": "informational",
+    "techniques": [
+      "tokenizer",
+      "control"
+    ],
+    "transformations": []
+  },
+  "corpus_id": "LF-CH-TOKX-0014",
+  "description": "Tokenizer security words control.",
+  "expected_rules": [],
+  "generator": "challenges.generate_challenge",
+  "ground_truth": "synthetic-generated",
+  "must_not_rules": [],
+  "repo_name": "tokenizer-security-words-control",
+  "repo_type": "model",
+  "safe_fixture": true,
+  "safety": "inert/canary-only",
+  "schema": 1,
+  "validation": {
+    "kind": "verify-package",
+    "target": "."
+  }
+}

README.md

@@ -0,0 +1,73 @@
+---
+license: apache-2.0
+tags:
+- layerfault
+- security-research
+- model-security
+- synthetic
+- adversarial-testing
+extra_gated_prompt: >-
+  This repository is a synthetic security-test artifact from the Layerfault
+  corpus. It intentionally contains adversarial characteristics (e.g.
+  suspicious pickle opcodes, executable-format smuggling, prompt-injection
+  strings) designed to exercise security scanner detection rules. It is
+  **not** a usable ML model and must never be loaded or executed outside an
+  isolated scanner-testing environment. By accepting, you confirm you
+  understand this repository is a test fixture, not production model
+  weights.
+extra_gated_button_content: I understand this is a security test fixture and accept the risk
+gated: auto
+---
+
+# tokenizer-security-words-control
+
+> **SECURITY TEST ARTIFACT: DO NOT USE AS A PRODUCTION MODEL**
+
+This repository is part of the Layerfault synthetic security corpus.
+It is deliberately constructed to contain security-relevant characteristics for scanner testing.
+
+**Corpus ID:** `LF-CH-TOKX-0014`
+
+## Purpose
+
+Tokenizer security words control.
+
+## Direct expected Layerfault rules
+
+- None; this repository is a control/comparison input.
+
+## Candidate rules
+
+These are deliberately plausible targets that remain marked as candidates until the exact
+Layerfault build used for certification confirms them.
+
+- None
+
+## Negative-control rules
+
+These should remain silent for this corpus item.
+
+- None
+
+## Safety
+
+The corpus uses fake secrets, loopback/`.invalid` network destinations, harmless marker output,
+and synthetic model behavior only. It is intended for static scanning and isolated security testing.
+
+## Challenge classification
+
+- Severity: **informational**
+- Difficulty: **compound**
+- Expected admission decision: **PASS**
+- Control type: **negative**
+- Attack surface: tokenizer-processor
+- Techniques: tokenizer, control
+- Transformations: none
+
+### Ground-truth oracle IDs
+
+- `LF-ORACLE-TOKX-0014`
+
+These oracle IDs describe synthetic ground truth. They do not claim that a
+matching Layerfault detector already exists. A challenge may intentionally expose
+a scanner blind spot and remain unmapped until the detector is implemented.

SECURITY.md

@@ -0,0 +1 @@
+Documentation only: hidden prompt role remap introspection dynamic include.

SHA256SUMS.json

@@ -0,0 +1,32 @@
+[
+  {
+    "path": "CHALLENGE_ORACLE.json",
+    "sha256": "a262a1eff2e885ee965243086b576570fb8d1bab12ddc72a37f1a9eaee93b47d",
+    "size": 611
+  },
+  {
+    "path": "FIXTURE_LAYOUT.json",
+    "sha256": "ac2a33c2038979c4c57f4f66f9c5340ee93b9147df683dbec28d61a9cb008950",
+    "size": 238
+  },
+  {
+    "path": "LAYERFAULT_CORPUS.json",
+    "sha256": "aa3244e33a56da7a2a0430436c1b6bfa578ef322970b26c2831fb25c7df92367",
+    "size": 848
+  },
+  {
+    "path": "README.md",
+    "sha256": "5cbddd8ef7375d91eb80cbce02181bedb69b0f23401529704ec36741a69e67d4",
+    "size": 2213
+  },
+  {
+    "path": "SECURITY.md",
+    "sha256": "ec33d010280df14c7e4146d67662ab296e50301e9bab87a1fa0345c47bb9ef3c",
+    "size": 76
+  },
+  {
+    "path": "tokenizer_config.json",
+    "sha256": "d3b9f2147b52291c7821c7c54b07c410bda9b378151f59af1bd3be83f699952f",
+    "size": 214
+  }
+]

tokenizer_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<bos>",
+  "chat_template": "{% for message in messages %}{{ message[\"role\"] }}: {{ message[\"content\"] }}{% endfor %}",
+  "eos_token": "<eos>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}