Make AI coding agents architecture-aware: baseline-first, evidence-verified, drift-checked, and safe across long tasks.

Evidence-driven skill evolution for Hermes Agent — reports, dry-run proposals, candidate search, and guarded apply

Chrome-first shopping extension product family with 8 storefront apps, 1 Suite shell, a reviewer-facing release shelf, and a read-only stdio MCP surface.

Open-source AI-powered vulnerability detection with evidence, ownership, and verified closure.

随叫随用、证据链驱动的 CUMCM / 数学建模 Codex Skills：覆盖题面理解、建模、复现、图表、审查与提交门禁。

Archived TaijiOS legacy prototype; current main repo is yangfei222666-9/taiji.

福尔摩斯式证据驱动渗透测试 — Claude Code SKILL + Python 推理引擎。五阶段科学侦查方法论 + 三角色执行 + 贝叶斯推理 + 自进化知识库。

EDPA — Evidence-Driven Proportional Allocation. Derive hours from Git evidence. No timesheets.

talk-html — a Claude Code skill that turns agent answers into shareable HTML. 用 HTML 说话，不要用聊天滚动条。