
This Skill is built on practical expertise in AI agent safety, prompt-injection defense, source-to-sink privacy review, tool and MCP metadata integrity, memory provenance checks, autonomous action scoping, and evidence-based verification. It helps agents separate instructions from evidence before tool calls, uploads, publishing, deletion, code execution, or network output.

Audit and manage local AI Skills across Codex, Claude Code, and project workspaces. See what is actually active, detect duplicates and variants, compare changes over time, and generate an offline interactive management map. Read-only by default; every real change requires your approval.

This Skill is built on practical expertise in long-running AI workflow control, constraint tracking, handoff reliability, completion evidence, and instruction-change recovery. It helps agents keep the user's first, middle, and latest requirements aligned across multi-step work so older assumptions do not override newer directions and completion is backed by observable evidence.

Based on an MIT-licensed open-source agent skill. Unofficial and not affiliated with ASD or STEMG.

Advanced AI-powered football analysis combining real match statistics, mathematical modeling, tactical intelligence, and probabilistic predictions to evaluate teams, analyze performance, estimate match outcomes, and generate five possible score scenarios with transparent data sources and uncertainty assessment.

用简单的话说出你希望AI编码智能体(Claude Code、Codex、OpenClaw)稳定完成的任务,即可获得一套完整、可直接发布的Skill包:结构规范、带激活优化描述的SKILL.md,可选的配套子智能体定义与hooks片段,一份验证清单,以及大多数作者会跳过的部分——发布前的安全自查,用来揪出密钥泄露、权限过宽和提示词注入风险。这是一套可复用的创作流程,不是固定模板。无需API密钥,不涉及个人数据。