Research & Data

Codex skill for resolving and downloading public GEO, NCBI, and ENA data with IBM Aspera ascp

Pre-registered A/B/C benchmark: Superpowers vs OpenSpec vs baseline, on real brownfield ops code. Methodology frozen before runs; results published either way.

One persona, one frozen memory, N models: how much of an agent's character survives a model swap? Harness + blind multi-lens judge panel + cross-vendor rank control + contamination probe. Results included: 7 models, spread 4.80 -> 2.20.

Voice -> text -> your personal knowledge base. A primitive, not a platform: 4 small Python scripts turn voice notes into linked, tagged, searchable markdown.

CharmOS / C(H+A)RM — open-source framework to manage humans AND AI agents as first-class relationships, on a personal second brain, driven by the RDR loop (Recall to Deep Research). Manifest-first, Apache-2.0.

Claude Desktop on Windows will not reopen after an update and you reboot. A dead instance is holding the single-instance lock. This kills only the stale processes, relaunches, and snapshots the failing state so the bug can be reported with artifacts. One PowerShell file, MIT.

Search local Agent Skill collections without loading them all.

Makes your AI agent answer like a briefing to a busy, skeptical boss. Conclusion first, no padding, no agreement it doesn't mean.

Open up your internal work without leaking it: substitute personal data with plausible fakes of the same shape (not <REDACTED>), then gate the WHOLE tree before the push. Stdlib-only, with the eleven traps it cost us.

Geography literature review research skill: distills HOW top journals structure reviews (Book-to-Skill), then runs an evidence-first multi-agent pipeline — Google Scholar API-only discovery, human-gated full-text acquisition, Zotero-verified citations, draft figures, independent audit.

苏格拉底式「反向提问」法 · Agent Skill：AI 反过来提问，引导用户自己理清思路、找到卡点、产出行动规划

Agent Skills: agent-thread-visualizer — execution maps / session timelines for Cursor, Codex, Claude Code, Workbuddy, kimi-code

让飞书 Aily 与 WorkBuddy 通过自然语言安全读取 GA4 的双平台 Skill 与 Reader API

Un agente piensa, otro construye y otro intenta romper lo construido. Así, la IA no reemplaza el proceso de ingeniería: lo estructura, lo acelera y lo vuelve más verificable.

Evidence-driven tutoring skills for Claude Code and Codex.

A personal stack of agent skills for reviewable planning, disciplined execution, and durable engineering lessons.

For You Agent——AI 时代的个人随身数字人格。把你的模型、AI 账号、技能、提示词和工作方式，带到每一个 AI 工具里。

Legal OA literature discovery, download, reporting, and local Zotero import Codex skill.

Skill-aware fixture diff reports for agent skill regression checks

A Codex skill for software-sector fund flows, industry analysis, and company deep dives.

一个面向数学、量化金融、机器学习和深度学习科研的可审计、自我迭代 Agent 框架。系统通过记录实验过程与失败原因、提炼可复用启发式规则、生成回归测试，并利用独立 Evaluator 和版本晋级机制持续优化研究 Skill。

Codex skill for professional literature search, Zotero deduplication, DOI import, project tagging, and verification reports.

Codex skill for GA4, GSC, Google Ads, and Shopify DTC growth reports

Agent skill that builds Open Knowledge Format (OKF v0.2) bundles — extracts metadata from CSVs, databases & BigQuery into markdown+YAML concept docs, enriches them from web documentation, and generates an interactive knowledge-graph visualization.