Update model card: v11 metrics + Further Reading section with blog post link

README.md

@@ -29,19 +29,19 @@ model-index:
           split: held_out_eval
         metrics:
           - type: f1
-            value: 0.8448
+            value: 0.926
             name: Macro F1
           - type: f1
-            value: 0.7857
+            value: 0.9007
             name: Injection F1
           - type: f1
-            value: 0.9040
+            value: 0.9513
             name: Benign F1
           - type: precision
-            value: 0.9362
+            value: 0.8608
             name: Injection Precision
           - type: recall
-            value: 0.6769
+            value: 0.9444
             name: Injection Recall
 ---
 
@@ -221,10 +221,25 @@ If you use this model in research, please cite:
 
 ---
 
-## Related Resources
-
+### Related Resources
 - [SkillScan project website](https://skillscan.sh)
 - [skillscan-security (rules, scanner, CLI)](https://github.com/kurtpayne/skillscan-security)
 - [Base model: protectai/deberta-v3-base-prompt-injection-v2](https://huggingface.co/protectai/deberta-v3-base-prompt-injection-v2)
 - [ProtectAI/rebuff — prompt injection detection research](https://github.com/protectai/rebuff)
 - [OWASP Top 10 for LLM Applications — LLM01: Prompt Injection](https://owasp.org/www-project-top-10-for-large-language-model-applications/)
+
+---
+
+## Further Reading
+
+**[What Are AI Agent Skills, and Why Do They Need a Security Model?](https://skillscan.sh/blog/skills-security-model)**
+
+A technical explainer for security engineers and enterprise architects covering:
+
+- What skills are — runbooks for agentic consumption, not traditional code, but often shipping with code; why the distinction matters for security
+- Five real attack archetypes with sanitized examples: README-driven dropper (AMOS/NemoClaw pattern), telemetry exfiltration disguised as analytics, indirect injection via trusted data channels, hallucination squatting, and goal substitution via jailbreak framing
+- How static analysis catches each archetype before runtime — with the actual rule or ML finding shown for each example
+- Where this model fits in the broader security stack: what it covers, what requires dynamic analysis (`skillscan-trace`), and what requires infrastructure controls (egress filtering, DNS-layer blocking)
+- Recommended enterprise posture: CI/CD gate setup, ML detection for high-risk skill directories, pre-production trace review, and infrastructure backstop
+
+The blog post uses the same five archetypes that are represented in this model's held-out eval set, making it a useful companion for understanding what the model is trained to detect and where its boundaries are.