You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
<h3>SkCC: The Final Form — Paper Submission & Open Source</h3>
60
-
<p>Unifying the project under the SkCC name, completing the NeurIPS 2026 paper, and releasing the full open-source compiler with benchmark leaderboard.</p>
72
+
<p>Unifying the project under the SkCC name, completing the paper, and releasing the full open-source compiler with benchmark leaderboard.</p>
<li>This website with benchmark leaderboard and development blog</li>
92
92
</ul>
93
93
94
-
<p>The paper was submitted to AgentSkills'26 (workshop at NeurIPS 2026) and uploaded to arXiv (2605.03353).</p>
94
+
<p>The paper was submitted to the AgentSkills'26 Workshop at ACM CAIS 2026 and uploaded to arXiv (2605.03353).</p>
95
95
96
96
<h2>What's Next</h2>
97
97
<p>SkCC is just the beginning. The compiler architecture is designed for extensibility — new framework emitters can be added by implementing a single trait. We're exploring:</p>
<p>On May 12, 2026, SkCC was featured in a full detailed article by <strong>小源学AI</strong> (Xiaoyuan Learns AI), a prominent AI-focused media account on Xiaohongshu (RED). The article provided an in-depth introduction to SkCC's core concepts — explaining how the compiler transforms a single SKILL.md source into framework-native formats for Claude Code, Codex, Gemini CLI, and Kimi CLI.</p>
<p>The coverage highlighted SkCC's key innovations: the four-phase compilation pipeline, the SkIR intermediate representation that decouples skill semantics from formatting, and the Anti-Skill Injection mechanism that provides compile-time security hardening. The article resonated strongly with the AI developer community, garnering 10 likes and 25 saves — indicating significant interest and bookmarking for future reference.</p>
68
+
69
+
<p>This marked SkCC's first appearance in Chinese-language AI media, helping bridge the gap between the academic research community and the broader developer ecosystem.</p>
70
+
71
+
<ahref="https://www.xiaohongshu.com/discovery/item/6a0218ef000000000602113c?source=webshare&xhsshare=pc_web&xsec_token=ABCgPWQk6960o0BUCUSfgYCZdMPVuPH3rS5HZgM40vgyo=&xsec_source=pc_share" target="_blank" class="media-link">🔗 View on Xiaohongshu (RED)</a>
72
+
</div>
73
+
</div>
74
+
</section>
75
+
76
+
<footer>
77
+
<divclass="footer-inner">
78
+
<divclass="footer-brand"><h3>SkCC</h3><p>Portable and Secure Skill Compilation for Cross-Framework LLM Agents.</p></div>
<p>On May 14, 2026, SkCC was featured by <strong>每日ComputerScience</strong> (Daily Computer Science), a respected computer science media account on Xiaohongshu (RED). The article, titled "中大新方法让Skill一次编译适配多个模型" (SYSU's New Method Enables One-Time Skill Compilation Across Multiple Models), highlighted SkCC's practical value for the AI agent ecosystem.</p>
<p>The coverage emphasized SkCC's origin at Sun Yat-sen University (SYSU) and its core value proposition: write a skill once, compile it to run optimally on Claude Code, Codex, Gemini CLI, and Kimi CLI. The article explained how SkCC's compiler-based approach reduces the O(m×n) adaptation burden to O(m+n), making cross-framework skill deployment practical for the first time.</p>
68
+
69
+
<p>The post received strong engagement with 32 likes and 23 saves, reflecting growing interest from the Chinese AI community in systematic solutions for agent skill portability and security.</p>
70
+
71
+
<ahref="https://www.xiaohongshu.com/discovery/item/69fc04f3000000003502aae8?source=webshare&xhsshare=pc_web&xsec_token=ABqP8sZrZv0GMo41cMl61zLNWusMK7d90P50ScYa5ucxU=&xsec_source=pc_share" target="_blank" class="media-link">🔗 View on Xiaohongshu (RED)</a>
72
+
</div>
73
+
</div>
74
+
</section>
75
+
76
+
<footer>
77
+
<divclass="footer-inner">
78
+
<divclass="footer-brand"><h3>SkCC</h3><p>Portable and Secure Skill Compilation for Cross-Framework LLM Agents.</p></div>
<h1><spanclass="gradient-text">Compilation for Skills:</span><br>Capable, Portable, and Securable</h1>
93
93
<pclass="hero-subtitle">SkCC is a compiler for LLM agents. Author skills once as SKILL.md, compile to Claude Code, Codex, Gemini CLI & Kimi CLI — with built-in security hardening.</p>
0 commit comments