Research & Data

机构级个股投研报告 Agent Skill（九章结构 + 多方法估值交叉验证），适用于 Claude Code / Cowork / Codex 等 AI Agent 工具

Defensive security research: MCP/web findings workflow + Agent Skills supply-chain scanner

A Codex skill for auditing and reconciling archive inventories across local, cloud, web, database, and media sources—without turning uncertainty into false missing-file claims.

A Codex skill for structured Arabic textbook learning with full vocalization, hint-first exercises, correction, and review.

Wenyu privacy-aware Bazi analysis engine and Codex skill

Codex skill for safely developing a PDD ecommerce AI analyst browser extension and local analysis service

Codex skill for staged, auditable batch flow cytometry analysis using FlowJo, FlowKit, compensation QC, reviewed gating, PDF reports, and Excel exports.

Local-only Codex skill for decrypting and querying an authorized WeChat for Mac database

Codex skill for evidence-scored Chinese AI intelligence daily HTML reports

Agent skill: EU Cyber Resilience Act vulnerability & reporting obligations, cited live via the Ansvar Gateway MCP connector

食品标签判断器 · 拿一张中国预包装食品标签，读懂它 —— 国标为锚、三态判断、认知归人。一个给 AI Agent 用的 Skill。

中国大陆版印象笔记（Yinxiang Biji / YXBJ / Evernote China）安全自动化 CLI 与 Claude Code Plugin

Shepherd is a Rust-native development platform that supervises AI coding agents (Claude Code, Codex) and verifies requirement completeness—》HumanInTheLoop

让 LLM agent 高效读懂任何文档：PDF/Word/PPT/Excel 转 Markdown，PDF 智能分流（文本页 PyMuPDF + 公式页/扫描页视觉 OCR）

Universal data-source assistant. Use this skill when the user wants external structured data such as stocks, financial reports, technical indicators, A-share/HK/US markets, global macroeconomics, Chinese enterprise registry information, arXiv papers, Google Scholar results, or Chinese laws/regulations and judicial cases. This plugin exposes tools via MCP server `plugin-kimi-datasource_data`; call them in the flow `mcp__plugin-kimi-datasource_data__get_data_source_desc` → `mcp__plugin-kimi-datasource_data__call_data_source_tool`.

Apply a senior-analyst reasoning discipline: reconstruct the real request, find the load-bearing assumption, verify against the actual artifact instead of pattern-matching, keep uncertainty visible, lead with the conclusion, and try to break the answer before delivering it. Use for ANY non-trivial task — analysis, debugging, planning, decisions, code, math, research, review, or writing that must be right — and especially when the user says "think carefully," "double-check," "be rigorous," or "are you sure," or when they will act on the answer without verifying it themselves. If being wrong would cost the user real time, money, or credibility, apply this skill even if they didn't ask.

Debug like Fable 5 — systematic root cause analysis with natural reasoning flow. Use this skill when you encounter an error, unexpected behavior, failing tests, or anything that does not work as intended.

Use when implementing, integrating, upgrading, debugging, or answering anything involving third-party APIs, libraries, frameworks, CLIs, cloud services, model/provider SDKs, fast-moving product behavior, user requests for latest/current/official behavior, unfamiliar repo docs/specs, errors that may indicate API drift, or high-stakes auth, security, billing, data, migration, deployment, compliance, or privacy behavior. Forces Codex to web-search for current official docs and read primary docs before assuming from memory.

Use when running Claude Fable on codebase-heavy or token-heavy work and the user wants Fable to orchestrate research, coding, and testing while cheaper subagents do bounded heavy lifting.

Apply the same orchestration as `/efficient-fable` to any high-cost frontier model: delegate research, coding, and testing to cheaper subagents while keeping planning, synthesis, and final review with the expensive model.

Codex Skill for scope-controlled security penetration testing, vulnerability scanning, evidence capture, reporting, and retesting.

Evidence-aware literature research workflow and reusable Codex skills for economics, finance, and management research

A security scanner for Agent Skill packages. Skill Sentinel uses multi-agent AI analysis to detect prompt injection, data exfiltration, command injection, malware, and other threats hiding in skill packages for Cursor, Claude Code, Codex, and OpenClaw.

Citation-first Codex Skill for official sources, public X conversations, and current social research.