Publish Layerfault synthetic corpus LF-CORPUS-HEUR-0001

LAYERFAULT_CORPUS.json

@@ -0,0 +1,19 @@
+{
+  "candidate_rules": [],
+  "corpus_id": "LF-CORPUS-HEUR-0001",
+  "description": "GGUF chat-template metadata containing a base64-encoded synthetic instruction-override phrase.",
+  "expected_rules": [
+    "LF-HEUR-DECODED-MATCH"
+  ],
+  "generator": "security_edges.decoded_heuristic",
+  "ground_truth": "synthetic-generated",
+  "must_not_rules": [],
+  "repo_name": "decoded-hidden-instruction",
+  "repo_type": "model",
+  "safety": "inert/canary-only",
+  "schema": 1,
+  "validation": {
+    "kind": "inspect",
+    "target": "model.gguf"
+  }
+}

README.md

@@ -0,0 +1,55 @@
+---
+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
+---
+
+# decoded-hidden-instruction
+
+> **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-CORPUS-HEUR-0001`
+
+## Purpose
+
+GGUF chat-template metadata containing a base64-encoded synthetic instruction-override phrase.
+
+## Direct expected Layerfault rules
+
+- `LF-HEUR-DECODED-MATCH`
+
+## 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.

SHA256SUMS.json

@@ -0,0 +1,22 @@
+[
+  {
+    "path": "LAYERFAULT_CORPUS.json",
+    "sha256": "52d31e5609ac2cd63d5922ef2f5c727a81266c326d616c24e7d119112edafff7",
+    "size": 542
+  },
+  {
+    "path": "README.md",
+    "sha256": "8691a02c19f218e1e38502be7a64d4d5839a9b6ab564ec559b851813b18d2f34",
+    "size": 1707
+  },
+  {
+    "path": "model.gguf",
+    "sha256": "a29bfbe9d715d7d956dc42568340d1572d1b5787c7cdd85de05a90c3879335ce",
+    "size": 160
+  },
+  {
+    "path": "security-note.txt",
+    "sha256": "a8e06aefb5f54b7f104a826ac07d1fd8b0850a824878293fbac9e2ca7645cf76",
+    "size": 137
+  }
+]

security-note.txt

@@ -0,0 +1 @@
+This repository's security-relevant text lives inside model.gguf's tokenizer.chat_template metadata field, not this file. See README.md.