Taxonomies
Every Keryn rule names the frameworks it relates to. These pages turn that around: pick an entry and see the rules that cover it, how sure they are, and the edges they inspect.
This is an edge map, not a score. A count says how many rules look at a class of weakness and how sure they are, not how safe a system is; a class with few rules may simply need few, and no number here means a framework is satisfied. Confidence: P0 exploitable as written, P1 dangerous in context, P2 heuristic or posture. Edges are the relationships the rules inspect (reaches, calls, reads, writes, sends, grants, depends_on, trusts).
| Framework | Entries | Rules mapped | Notes |
|---|---|---|---|
| OWASP Top 10 for LLM Applications 2025 | 10 | 1438 | ten entries, from prompt injection to unbounded consumption |
| OWASP Top 10 for Agentic Applications | 10 | 1134 | ten entries for agents that plan and use tools |
| CWE | 143 | 1412 | 143 weaknesses, the 30 most covered in detail |
| NIST AI RMF 1.0 | 4 | 95 | the four core functions |
The mappings are references for readers who work in those frameworks. None is a compliance or conformance claim. Rules are listed in the rule catalog, which you can also filter by framework entry.