Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions data/scan-corpus-failures.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
{
"_description": "Failed scan attempts logged by the weekly corpus refresh routine.",
"failures": [
{
"timestamp": "2026-08-30T19:35:03Z",
"repo": "szybnev/DVLA",
"reason": "HTTP 400 from scan-public API"
}
]
}
26 changes: 22 additions & 4 deletions data/scan-corpus.json
Original file line number Diff line number Diff line change
Expand Up @@ -3,9 +3,9 @@
"_description": "Public scan corpus — the live dataset behind State of AI Agent Security 2026. Refreshed weekly by an automated routine.",
"_methodology": "See data/scan-corpus-readme.md for methodology, included repo criteria, and rotation cadence.",
"aggregates": {
"total_scans": 2,
"total_findings_observed": 9,
"average_governance_score": 16.0,
"total_scans": 3,
"total_findings_observed": 11,
"average_governance_score": 19.67,
"finding_categories_aggregated": [
"sql_injection",
"injection",
Expand All @@ -15,7 +15,7 @@
"resource_exhaustion",
"code_injection"
],
"last_refreshed": "2026-05-10T19:35:03Z",
"last_refreshed": "2026-08-30T19:35:14Z",
"first_scan_date": "2026-05-04T12:10:06Z"
},
"scans": [
Expand Down Expand Up @@ -56,6 +56,24 @@
"code_injection",
"governance"
]
},
{
"timestamp": "2026-08-30T19:35:14Z",
"repo": "merlvintan/damn-vulnerable-llm-agent",
"report_id": "c1253198-8871-4a0c-8a95-b5507672a32c",
"findings_count": 2,
"critical_count": 0,
"high_count": 2,
"medium_count": 0,
"low_count": 0,
"governance_score": 27,
"risk_score": 22,
"eu_ai_act_readiness": "PARTIAL",
"top_finding_categories": [
"sql_injection",
"injection"
],
"frameworks_covered": ["langchain"]

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P3: The new scan entry introduces a per-scan frameworks_covered field that no other corpus entry has and that data/scan-corpus-readme.md does not document in the entry structure list. Data consumers parsing scan-corpus.json now must handle a field present in one entry but absent in the other two, and its semantics are undocumented. Either add it to all entries consistently and document it in the readme, or drop it from this entry so the schema stays uniform.

Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At data/scan-corpus.json, line 76:

<comment>The new scan entry introduces a per-scan `frameworks_covered` field that no other corpus entry has and that `data/scan-corpus-readme.md` does not document in the entry structure list. Data consumers parsing scan-corpus.json now must handle a field present in one entry but absent in the other two, and its semantics are undocumented. Either add it to all entries consistently and document it in the readme, or drop it from this entry so the schema stays uniform.</comment>

<file context>
@@ -56,6 +56,24 @@
+        "sql_injection",
+        "injection"
+      ],
+      "frameworks_covered": ["langchain"]
     }
   ]
</file context>

}
]
}
Loading