Skip to content

TOOLS.md

Latest commit

 

History

History
80 lines (57 loc) · 5.78 KB

TOOLS.md

File metadata and controls

80 lines (57 loc) · 5.78 KB
Raw

Tools

This is a curated directory of Codex-adjacent resources for reviewable office work. Entries are grouped by work use, not ranked.

Each entry should have a public source, a real work problem, a review boundary, and a stated limit. See data/catalog.yml for the structured catalog.

Official OpenAI/Codex

Official resources are the default route for learning Codex behavior. Recheck the source before relying on product availability, access, pricing, or model details.

Entry What it helps with Review boundary
OpenAI Codex docs Understand Codex surfaces, configuration, workflows, MCP, skills, and plugins Treat docs as authoritative for OpenAI product behavior, but recheck volatile pages
Codex for Work Academy Learn official work-use examples and framing Do not infer support for every office workflow
openai/codex Inspect the open-source Codex CLI repository Repository status and releases can change
openai/skills Browse reusable Codex skills Skill fit depends on the specific task and review process

MCP And Connectors

MCP and connector entries belong here when they help Codex-style agents reach a work system or source. Prefer read-only use until the review and permission boundary is clear.

Entry What it helps with Review boundary
Codex MCP docs Configure MCP servers for Codex Recheck current config behavior before editing local setup
Model Context Protocol docs Understand MCP as a connector pattern MCP support differs by client and server
github/github-mcp-server Connect agents to GitHub repositories, issues, and pull requests Confirm repo permissions and write capabilities before use
microsoft/playwright-mcp Give agents browser automation through Playwright Browser automation can expose or change sensitive web state
ChromeDevTools/chrome-devtools-mcp Inspect and debug browser pages through Chrome DevTools Developer debugging context may include private page data
upstash/context7 Retrieve current developer documentation through MCP Verify retrieved docs against primary sources for critical work

Workspace-Native Agent Apps

This category covers apps and patterns that run close to existing work surfaces. Keep the language soft and independent; do not imply a category is official unless the source says so.

Entry What it helps with Review boundary
Workspace agents in ChatGPT Understand OpenAI's workspace-agent direction and examples Do not overstate availability beyond the source
OpenAI Apps SDK Build apps that extend ChatGPT with tools and UI App behavior, submission, and platform rules are source-sensitive

Document, Spreadsheet, And Deck Tools

Use this section for tools that help create, inspect, transform, or review office work products. Listings must say what source material the tool uses and what a human should review.

Entry What it helps with Review boundary
openai/skills Reusable Codex task bundles, including document-like workflows when a skill exists A skill still needs input review, output review, and source boundaries
Workflow starter set Copy-ready prompts for briefs, notes, memos, spreadsheets, decks, and specs These are templates, not guarantees

Review And Source-Trace Tools

These entries help make Codex output easier to verify. They are not substitutes for human judgment.

Entry What it helps with Review boundary
CHECKLISTS.md Human review, source traceability, MCP privacy, and approval checks Adapt to your team's risk and policy context
github/github-mcp-server Keep code and issue work connected to source records Confirm permissions before write-capable use
upstash/context7 Pull current docs into agent workflows Recheck important facts against the source doc

Developer And Agent Frameworks

Framework entries belong here when they help someone build Codex-adjacent tools, agents, workflows, or review loops.

Entry What it helps with Review boundary
OpenAI Agents SDK docs Understand OpenAI's agent-building framework Implementation details change; recheck docs
openai/openai-agents-js Build JavaScript or TypeScript agents Review releases, examples, and security posture before adoption
openai/openai-agents-python Build Python agents Review releases, examples, and security posture before adoption
openai/openai-cookbook Find examples and implementation patterns Examples are starting points, not production guarantees

Suggest A Tool

Use the tool submission form or open a PR. A useful tool submission includes:

  • the real work problem it solves
  • source URL
  • Codex relationship
  • output it helps produce
  • human review boundary
  • limits and access notes
  • submitter affiliation