Frontier World Skills

Reusable, production-oriented Codex Skills maintained by Frontier World in the current FrontierOpen GitHub organization.

This repository contains self-contained skill packages for repeatable research, content, publishing, and automation workflows. Each skill combines operating instructions with the scripts, references, and assets required to run and validate it.

Available skills

Skill	Description	Status
frontier-signals	Turns current AI and technology news into a sourced daily article and prepares consistent WeChat, web, and Feishu editions.	Active
daily-tech-brief	Collects and renders a ten-item daily technology briefing. Retained while scheduled workflows migrate to Frontier Signals.	Legacy

Repository structure

skills/
└── <skill-name>/
    ├── SKILL.md       # Skill definition and workflow
    ├── agents/        # Agent-facing metadata
    ├── scripts/       # Deterministic automation and tests
    ├── references/    # Policies, schemas, and integration notes
    └── assets/        # Templates and example inputs

Every top-level skill directory is designed to remain independently understandable, testable, and reusable.

Installation

Clone the repository:

git clone https://github.com/FrontierOpen/skills.git
cd skills

Install a skill by copying or linking its directory into the skills directory used by your Codex environment. For example:

ln -s /path/to/skills/frontier-signals /path/to/codex-home/skills/frontier-signals

Open the selected skill's SKILL.md for its prerequisites, workflow, and validation commands.

Development standards

• Keep one skill per top-level directory and match the directory name to the name field in SKILL.md.
• Make each skill self-contained, deterministic where possible, and independently testable.
• Keep reusable logic in scripts instead of embedding large command sequences in instructions.
• Document external services, schemas, safety constraints, and failure behavior in references/.
• Include example inputs only as structural fixtures; never present sample data as current production data.
• Do not commit credentials, authentication state, generated working files, or private user data.

Validation

Run the checks defined by the individual skill before opening a pull request. For frontier-signals:

cd frontier-signals
npm test

Contributing

Issues and pull requests are welcome. A contribution should include:

1. A focused use case and clear trigger conditions.
2. A complete SKILL.md with explicit inputs, outputs, and failure rules.
3. Tests for deterministic scripts and validation for generated artifacts.
4. Documentation for required permissions and external integrations.
5. No secrets, personal information, or environment-specific runtime data.

Please keep changes scoped to one skill whenever possible.

Security

Never commit API keys, access tokens, browser sessions, private documents, or production data. If a credential is exposed, revoke it immediately and report the incident through the repository's security channel rather than a public issue.