Help Center
Spotlight Skills

Which AI tools can use a business context pack?

Around 40 products support the open Agent Skills specification, including Claude, ChatGPT and Codex, GitHub Copilot, Cursor, Gemini CLI, and VS Code. One Markdown file works across all of them, so you are not locked into any single AI vendor.

Last updated

Roughly 40 products support the open Agent Skills specification, so one context pack works across the AI tools your team already uses. Notable ones include:

  • Claude (Claude Code and the Claude apps)
  • ChatGPT and Codex CLI
  • GitHub Copilot and VS Code
  • Cursor
  • Gemini CLI
  • JetBrains Junie, AWS Kiro, Block Goose, and others

Why it's portable

Anthropic published the Agent Skills specification on December 18, 2025 as an open standard rather than a proprietary feature. Adoption moved unusually fast — Microsoft added it to VS Code and OpenAI added it to ChatGPT and Codex within about 48 hours, and the ecosystem grew from there.

The practical consequence for a business: your context pack is not a bet on one AI vendor. It's a Markdown file. If your team switches from ChatGPT to Claude next year, or runs both, the same file comes along. That portability is why we build in this format rather than using any single tool's proprietary "custom instructions" box.

What about tools that don't support it?

Plenty of AI features inside other software — a CRM's built-in assistant, for example — have their own settings and won't read a skills file. Two options:

  • Paste the relevant sections into that tool's own instructions field. Less elegant, works fine.
  • Keep the pack as your source of truth and copy from it. This is the important habit: one authoritative version that other places derive from, rather than five drifting copies.

Does the model matter?

Less than you'd think for this purpose. A good context pack improves output on every reasonably capable model, because it's supplying missing information rather than compensating for weak reasoning. Use whichever model your team prefers — we run a three-model stack ourselves and the same pack serves all three.


Not sure what your team is actually using? That's part of what the free 45-minute consultation figures out. Book it here.

Related articles

First 4 calls freeNo obligation

Want us to check this for you?

Book a free strategy call and we'll audit your setup — indexing, robots.txt, schema — and tell you exactly what's blocking you.

Typical response: 24–48 hours. Same day for active engagements.

Didn’t answer your question? Ask us directly — we reply within 24–48 hours.