Catch-up — Agent Skills Move to a Neutral Home: Write Yours to the Standard
A repeatable workflow is more portable when its instructions and resources live in a standard skill folder rather than a tool-specific prompt.
- Anthropic describes Agent Skills as an open standard for giving AI agents specialised knowledge and workflows and says it donated the standard to AAIF.
- Anthropic says more than 70 agent platforms integrated Agent Skills after the standard was published in December 2025.
- Anthropic defines a skill as a folder with a SKILL.md that teaches a task or domain, plus any needed scripts or references.
- Anthropic says one million skills have been published and nearly 280 million installed.
Catch-up written 24 October 2026 for 2026-10-22; covers a source published 2026-10-22.
Portable workflow packages reduce the work of recreating team knowledge in every agent, while testing reveals where implementations still differ.
Choose a weekly workflow and write a skill containing its trigger, steps, inputs, example, and success check. Run the same skill in two agents.
Put it to work 30 minutes
A portable SKILL.md tested in two agents
Turn one weekly workflow into a SKILL.md: when to use it, steps, inputs, one example, and the check that proves it worked. Try two agents.
See a worked example
Use a real, bounded workflow and record unknowns rather than estimating them. Keep the result small enough to review with a teammate.
- Workflow or decision
- A real workflow the team repeats each week.
- Evidence to record
- Quality, time, cost, failures, and reviewer notes.
- Decision rule
- Keep the current approach unless the agreed checks pass.
Your turn
Name the exact case you will test.
List the observations needed to make the decision.
Write what the evidence will make you do.
Your draft stays in this browser. No account needed.
Anthropic's install and adoption figures are its own. Moving a standard to a foundation does not guarantee identical implementations, so test each agent you use.
Reported facts and quotations were checked against the named sources; the recommended exercise is editorial guidance to test in your own context.
Did the same skill complete the workflow correctly in both agents?
Watch: Agent-specific changes required by the skill · Failures against the written success checkKeep Going