Codex Academic Skills

English | 简体中文

A practical index of research-oriented skills that work with OpenAI Codex.

It focuses on official skills, repos that clearly say they support Codex, and open Agent Skills that Codex can read without much adaptation.

Each skill is listed once, under the workflow where it is most likely to be useful. Before installing or relying on one, check the upstream SKILL.md; this repo is a map, not the source of record.

---

Table of Contents

• What Are Codex Skills?
• Inclusion Rules
• How To Use This List
• Core Picks
• Skill List
• Installation and Usage
• License
• References

---

What Are Codex Skills?

Codex skills are folder-based instruction bundles that help Codex handle specific tasks more reliably. A typical skill includes a SKILL.md file and may also include scripts, templates, or references.

Codex first sees each skill's name, description, and path. It reads the full SKILL.md only when the task matches. That keeps context usage under control, but it also means large skill sets have a real cost. OpenAI's docs note that the initial skill list has a context budget; with many installed skills, descriptions may be shortened, and very large sets may omit some skills from the initial list.

---

Inclusion Rules

This list keeps entries that satisfy at least one of the following:

• official OpenAI Codex skills
• repositories that explicitly document Codex support or interoperability
• open Agent Skills that Codex can read directly or with minimal adaptation

This list intentionally excludes:

• skills that are exclusive to other platforms
• entries whose upstream links disappeared without a clear replacement path
• highly overlapping entries that do not add clear value for research workflows
• repositories whose Codex compatibility is unclear

---

How To Use This List

Treat this repository as a research-workflow index, not a marketplace. Start with the Core Picks below, then move into the category that matches your bottleneck.

Task-based scan:

• Research orchestration, project planning, and model choice: section 1.
• Context engineering, agent design, and structured outputs: section 2.
• Paper drafting, formal scholarly writing, and de-AI editing: section 3.
• Literature review, transcription, and evidence synthesis: section 4.
• Demos, figures, talks, and presentation assets: section 5.
• Data, retrieval, experiment tracking, and observability: section 6.
• Fine-tuning, evaluation, serving, and kernel optimization: section 7.
• Mechanistic interpretability and model analysis: section 8.
• Research artifacts, provenance, and reproducibility checks: section 9.

---

Core Picks

If you only want a small starter set, begin here. These skills cover broad research workflows, have clear upstream sources, and keep the active skill list from getting too crowded.

Skill	What It Does	Link
autoresearch	Orchestrates literature review, experiments, synthesis, and paper-writing workflows.	Orchestra-Research/AI-Research-SKILLs
notion-research-documentation	Researches across Notion and synthesizes cited briefs and reports.	openai/skills
openai-docs	Looks up current OpenAI product and API documentation with source-grounded guidance.	openai/skills
huggingface-best	Finds and compares candidate models using official Hugging Face benchmark leaderboards and metadata.	huggingface/skills
pdf	Reads, creates, and reviews PDFs when layout and rendering matter.	openai/skills
huggingface-papers	Looks up Hugging Face paper pages and structured paper metadata.	huggingface/skills
academic-plotting	Generates publication-quality charts, ablations, and architecture figures.	Orchestra-Research/AI-Research-SKILLs
jupyter-notebook	Creates clean, reproducible Jupyter notebooks for experiments and tutorials.	openai/skills
huggingface-datasets	Explores Dataset Viewer API metadata, rows, search, filters, parquet URLs, and statistics.	huggingface/skills
huggingface-local-models	Selects GGUF models, quantization levels, and llama.cpp commands for local inference.	huggingface/skills
hf-mem	Estimates inference memory and optional KV-cache requirements for Safetensors or GGUF models on the Hugging Face Hub.	huggingface/skills
huggingface-llm-trainer	Trains or fine-tunes language models with TRL on Hugging Face Jobs, including SFT, DPO, GRPO, and reward models.	huggingface/skills
lm-evaluation-harness	Runs standardized LLM benchmarks such as MMLU, HumanEval, GSM8K, and TruthfulQA.	Orchestra-Research/AI-Research-SKILLs
vllm	Serves LLMs with high-throughput inference and OpenAI-compatible endpoints.	Orchestra-Research/AI-Research-SKILLs
transformer-lens	Supports mechanistic interpretability with HookPoints, activation caching, and causal tracing.	Orchestra-Research/AI-Research-SKILLs
remove-ai-flavor	Revises Chinese prose to remove template-like AI phrasing while preserving facts, stance, and author voice.	B1lli/remove-ai-flavor-writing-skill
ara-compiler	Compiles papers, repos, logs, and notes into Agent-Native Research Artifacts with claims, evidence, and provenance.	Orchestra-Research/AI-Research-SKILLs

---

Skill List

1. Planning and Workflow

Skill	What It Does	Link
project-development	Scopes LLM projects and designs practical research-agent architectures.	muratcankoylan/Agent-Skills-for-Context-Engineering
notion-research-documentation	Researches across Notion and synthesizes cited briefs and reports.	openai/skills
notion-knowledge-capture	Captures conversations, notes, and decisions into structured Notion pages.	openai/skills
notion-meeting-intelligence	Prepares agendas, pre-reads, and decision docs in Notion.	openai/skills
autoresearch	Orchestrates literature review, experiments, synthesis, and paper-writing workflows.	Orchestra-Research/AI-Research-SKILLs
brainstorming-research-ideas	Guides structured ideation for high-impact research directions.	Orchestra-Research/AI-Research-SKILLs
creative-thinking-for-research	Applies creativity frameworks to generate less obvious research ideas.	Orchestra-Research/AI-Research-SKILLs
openai-docs	Looks up current OpenAI product and API documentation with source-grounded guidance.	openai/skills
define-goal	Turns fuzzy intentions into concrete, measurable objectives before long-running agent work.	openai/skills
huggingface-best	Finds and compares candidate models using official Hugging Face benchmark leaderboards and metadata.	huggingface/skills

2. Context and Agent Design

Skill	What It Does	Link
context-fundamentals	Explains how context works in agent systems.	muratcankoylan/Agent-Skills-for-Context-Engineering
context-degradation	Diagnoses lost-in-the-middle and other context failure modes.	muratcankoylan/Agent-Skills-for-Context-Engineering
context-compression	Compresses long sessions while preserving critical state.	muratcankoylan/Agent-Skills-for-Context-Engineering
context-optimization	Applies caching, masking, compaction, and partitioning to extend effective context capacity.	muratcankoylan/Agent-Skills-for-Context-Engineering
filesystem-context	Uses files as an overflow layer for scratch pads, shared agent memory, and just-in-time context loading.	muratcankoylan/Agent-Skills-for-Context-Engineering
latent-briefing	Studies KV-cache style latent handoff for reducing token cost in orchestrator-worker systems.	muratcankoylan/Agent-Skills-for-Context-Engineering
evaluation	Builds evaluation frameworks for agent systems with rubrics, regressions, and quality gates.	muratcankoylan/Agent-Skills-for-Context-Engineering
advanced-evaluation	Covers LLM-as-a-judge and bias-aware automated evaluation.	muratcankoylan/Agent-Skills-for-Context-Engineering
multi-agent-patterns	Designs supervisor, swarm, and hierarchical multi-agent systems with clearer coordination.	muratcankoylan/Agent-Skills-for-Context-Engineering
memory-systems	Designs persistent agent memory and evaluates retrieval quality across sessions.	muratcankoylan/Agent-Skills-for-Context-Engineering
tool-design	Designs agent-facing tools and MCP interfaces with clearer contracts and less ambiguity.	muratcankoylan/Agent-Skills-for-Context-Engineering
harness-engineering	Designs autonomous agent harnesses with locked metrics, editable surfaces, durable logs, rollback, and approval boundaries.	muratcankoylan/Agent-Skills-for-Context-Engineering
dspy	Uses declarative prompt programming and optimizers for structured agent workflows.	Orchestra-Research/AI-Research-SKILLs
guidance	Controls generation with regexes and grammars for structured outputs and multi-step prompting.	Orchestra-Research/AI-Research-SKILLs
instructor	Produces Pydantic-validated structured outputs for extraction, labeling, and automation.	Orchestra-Research/AI-Research-SKILLs
outlines	Constrains generation with grammars and finite-state machines for structured outputs.	Orchestra-Research/AI-Research-SKILLs

3. Writing and Scholarly Communication

Skill	What It Does	Link
pdf	Reads, creates, and reviews PDFs when layout and rendering matter.	openai/skills
huggingface-paper-publisher	Publishes papers on Hugging Face Hub and manages paper metadata.	huggingface/skills
ml-paper-writing	Writes publication-oriented ML, AI, and systems papers.	Orchestra-Research/AI-Research-SKILLs
systems-paper-writing	Provides paragraph-level blueprints for OSDI, SOSP, ASPLOS, NSDI, and EuroSys papers.	Orchestra-Research/AI-Research-SKILLs
remove-ai-flavor	Revises Chinese prose to remove template-like AI phrasing while preserving facts, stance, and author voice.	B1lli/remove-ai-flavor-writing-skill
unslop	Audits and rewrites AI-generated prose with fact preservation, voice presets, and explicit anti-pattern checks.	theclaymethod/unslop
peer-review	Produces structured manuscript or grant reviews with methodology, statistics, reproducibility, and reporting-standard checks.	K-Dense-AI/scientific-agent-skills

4. Literature Reading and Evidence Synthesis

Skill	What It Does	Link
transcribe	Transcribes interviews, meetings, or recorded talks with optional speaker diarization.	openai/skills
whisper	Runs multilingual speech recognition and translation for interviews, lectures, podcasts, and audio corpora.	Orchestra-Research/AI-Research-SKILLs
huggingface-papers	Looks up Hugging Face paper pages and structured paper metadata.	huggingface/skills
citation-management	Searches, verifies, deduplicates, and formats academic citations and BibTeX metadata.	K-Dense-AI/scientific-agent-skills
llamaindex	Builds document ingestion and retrieval pipelines for research corpora.	Orchestra-Research/AI-Research-SKILLs
faiss	Provides high-performance dense retrieval for paper collections.	Orchestra-Research/AI-Research-SKILLs
sentence-transformers	Generates embeddings for literature search, clustering, and retrieval.	Orchestra-Research/AI-Research-SKILLs

5. Visualization and Presentation

Skill	What It Does	Link
huggingface-gradio	Builds Gradio web UIs and interactive research demos in Python.	huggingface/skills
huggingface-spaces	Builds, deploys, and debugs Hugging Face Spaces across Gradio, Docker, Static, ZeroGPU, and paid hardware.	huggingface/skills
huggingface-trackio	Tracks training metrics, alerts, and dashboards with Hugging Face Trackio.	huggingface/skills
academic-plotting	Generates publication-quality charts, ablations, and architecture figures.	Orchestra-Research/AI-Research-SKILLs
presenting-conference-talks	Turns papers into Beamer or PPTX talk decks with speaker notes and scripts.	Orchestra-Research/AI-Research-SKILLs
speech	Generates narration, accessibility reads, and voiceovers via the OpenAI Audio API.	openai/skills
imagegen	Creates or edits bitmap figures, mockups, infographics, and visual assets.	openai/skills
transformers-js	Runs Hugging Face models in JavaScript for browser-side demos and interactive artifacts.	huggingface/skills
stable-diffusion	Generates figures, concept art, and presentation assets for multimodal research.	Orchestra-Research/AI-Research-SKILLs
scientific-visualization	Creates publication-ready scientific figures with journal styles, multi-panel layouts, and accessible palettes.	K-Dense-AI/scientific-agent-skills

6. Data, Retrieval, and Experiment Tracking

Skill	What It Does	Link
jupyter-notebook	Creates clean, reproducible Jupyter notebooks for experiments and tutorials.	openai/skills
hf-cli	Manages Hugging Face auth, repos, papers, datasets, buckets, jobs, and endpoints from the hf CLI.	huggingface/skills
huggingface-datasets	Explores Dataset Viewer API metadata, rows, search, filters, parquet URLs, and statistics.	huggingface/skills
huggingface-community-evals	Adds and manages evaluation results in model cards and custom HF Hub evaluations.	huggingface/skills
huggingface-tool-builder	Builds reusable scripts around the Hugging Face API for metadata collection and automation.	huggingface/skills
huggingface-local-models	Selects GGUF models, quantization levels, and llama.cpp commands for local inference.	huggingface/skills
hf-mem	Estimates inference memory and optional KV-cache requirements for Safetensors or GGUF models on the Hugging Face Hub.	huggingface/skills
ray-data	Scales batch inference, preprocessing, and multimodal ETL from one machine to clusters.	Orchestra-Research/AI-Research-SKILLs
nemo-curator	Curates training corpora with deduplication, quality filtering, PII redaction, and multimodal cleanup.	Orchestra-Research/AI-Research-SKILLs
weights-and-biases	Tracks experiments, sweeps, artifacts, and model registries.	Orchestra-Research/AI-Research-SKILLs
mlflow	Handles experiment tracking, model registry, deployment, and autologging workflows.	Orchestra-Research/AI-Research-SKILLs
tensorboard	Visualizes scalars, embeddings, profiles, and training diagnostics.	Orchestra-Research/AI-Research-SKILLs
langsmith	Adds tracing, evaluation, and monitoring to LLM research workflows.	Orchestra-Research/AI-Research-SKILLs
phoenix	Provides open-source tracing, evaluation, and experiment analysis.	Orchestra-Research/AI-Research-SKILLs

7. Training, Evaluation, and Serving

This category can grow quickly, so it keeps representative entries tied closely to research training, evaluation, or serving.

Skill	What It Does	Link
huggingface-llm-trainer	Trains or fine-tunes language models with TRL on Hugging Face Jobs, including SFT, DPO, GRPO, and reward models.	huggingface/skills
huggingface-vision-trainer	Trains detection and classification models with Transformers Trainer on Hugging Face Jobs or locally.	huggingface/skills
axolotl	Provides YAML-first fine-tuning workflows for LoRA, QLoRA, DPO, and multimodal training.	Orchestra-Research/AI-Research-SKILLs
llama-factory	Provides WebUI and CLI workflows for low-code fine-tuning across language and multimodal models.	Orchestra-Research/AI-Research-SKILLs
unsloth	Accelerates LoRA and QLoRA fine-tuning with lower memory use for local experiments.	Orchestra-Research/AI-Research-SKILLs
peft	Covers parameter-efficient fine-tuning with LoRA, QLoRA, DoRA, and adapters.	Orchestra-Research/AI-Research-SKILLs
trl-fine-tuning	Uses TRL for post-training workflows such as SFT, DPO, PPO, GRPO, and reward-model training.	Orchestra-Research/AI-Research-SKILLs
train-sentence-transformers	Trains bi-encoders, rerankers, and sparse encoders with sentence-transformers for retrieval and similarity workflows.	huggingface/skills
grpo-rl-training	Specializes in GRPO-based post-training for reasoning, verifiable tasks, and custom rewards.	Orchestra-Research/AI-Research-SKILLs
lm-evaluation-harness	Runs standardized LLM benchmarks such as MMLU, HumanEval, GSM8K, and TruthfulQA.	Orchestra-Research/AI-Research-SKILLs
bigcode-evaluation-harness	Benchmarks code models with HumanEval, MBPP, MultiPL-E, and pass@k workflows.	Orchestra-Research/AI-Research-SKILLs
nemo-evaluator	Runs reproducible multi-backend benchmarking across LLM and VLM benchmarks.	Orchestra-Research/AI-Research-SKILLs
vllm	Serves LLMs with high-throughput inference and OpenAI-compatible endpoints.	Orchestra-Research/AI-Research-SKILLs
sglang	Serves LLMs and VLMs with fast structured generation, prefix caching, and JSON/tool-calling workflows.	Orchestra-Research/AI-Research-SKILLs
llama-cpp	Runs quantized LLMs on CPUs, Apple Silicon, and non-CUDA hardware.	Orchestra-Research/AI-Research-SKILLs
cuda-kernels	Guides optimized CUDA kernel writing and benchmarking for Hugging Face diffusers and transformers.	huggingface/kernels
rocm-kernels	Guides optimized Triton kernel writing and benchmarking for AMD ROCm GPUs in diffusers and transformers workflows.	huggingface/kernels

8. Interpretability and Model Analysis

Skill	What It Does	Link
transformer-lens	Supports mechanistic interpretability with HookPoints, activation caching, and causal tracing.	Orchestra-Research/AI-Research-SKILLs
saelens	Trains and analyzes sparse autoencoders for monosemantic feature discovery and superposition research.	Orchestra-Research/AI-Research-SKILLs
nnsight	Runs local or remote interpretability experiments on PyTorch models, including very large models via NDIF.	Orchestra-Research/AI-Research-SKILLs
pyvene	Performs causal interventions, activation patching, and interchange intervention training on PyTorch models.	Orchestra-Research/AI-Research-SKILLs

9. Research Artifacts and Reproducibility

Skill	What It Does	Link
ara-compiler	Compiles papers, repos, logs, and notes into Agent-Native Research Artifacts with claims, evidence, and provenance.	Orchestra-Research/AI-Research-SKILLs
ara-research-manager	Records end-of-session research decisions, experiments, dead ends, and provenance into an ARA directory.	Orchestra-Research/AI-Research-SKILLs
ara-rigor-reviewer	Reviews Agent-Native Research Artifacts for evidence relevance, falsifiability, scope, coherence, and rigor.	Orchestra-Research/AI-Research-SKILLs

---

Installation and Usage

This repository is a curated list, not a package manager. Current Codex docs describe skills as the authoring format for reusable workflows; for broader distribution, package reusable skills as plugins.

Install a skill in Codex

Current Codex docs describe these standard skill locations:

• repository scope: .agents/skills/<skill-name>/
• user scope: ~/.agents/skills/<skill-name>/
• admin scope: /etc/codex/skills/<skill-name>/

Example 1: install an official curated skill from openai/skills

$skill-installer pdf

Example 2: install a Hugging Face skill manually

mkdir -p ~/.agents/skills
cd /tmp
git clone --depth 1 https://github.com/huggingface/skills.git
cp -R skills/skills/huggingface-papers ~/.agents/skills/

Example 3: install the Hugging Face Kernels Codex skill

kernels skills add --codex --global

Example 4: install the Orchestra Research bundle with its upstream installer

npx @orchestra-research/ai-research-skills

Use a skill in Codex

Once the folder is available in a valid Codex skill location, you can invoke it naturally in your prompt.

Examples:

• Use autoresearch to set up an experiment loop for this idea.
• Use huggingface-best to compare current open models for this benchmark.
• Use hf-mem to estimate whether this model fits on my GPU.
• Use academic-plotting to turn these ablation results into camera-ready figures.
• Use huggingface-local-models to choose a GGUF model for this laptop.
• Use remove-ai-flavor to clean up this Chinese project summary.
• Use ara-compiler to turn this paper and repo into a research artifact.

Recommended usage pattern

1. Install a small set of high-frequency skills first, then add project-specific ones as needed.
2. Pick one skill for one clear bottleneck.
3. Read the upstream SKILL.md before relying on the result.
4. If the source repository ships its own installer, plugin manifest, or fallback AGENTS.md, read its install docs before mixing methods.
5. For academic work, manually check citations, claims, equations, data handling, and benchmark settings.
6. If a skill touches remote services or external datasets, verify authentication, quotas, privacy, and licensing before running it at scale.

---

License

The content of this repository is released under the MIT License.

Third-party skills linked from this list keep their own licenses. Always check the original repository before installing or redistributing anything.

If you notice a dead link, a naming change, or a clearly better entry for the list, a short issue or PR is enough.

---

References

• Agent Skills – Codex
• openai/skills
• huggingface/skills
• Hugging Face Kernels skills add
• huggingface/kernels
• Orchestra-Research/AI-Research-SKILLs
• K-Dense-AI/scientific-agent-skills
• Agent-Skills-for-Context-Engineering
• B1lli/remove-ai-flavor-writing-skill
• theclaymethod/unslop
• Agent Skills Open Standard