Skip to content

Propose adopting Agent Skills#1165

Open
NullVoxPopuli wants to merge 5 commits intomainfrom
nvp/agent-skills
Open

Propose adopting Agent Skills#1165
NullVoxPopuli wants to merge 5 commits intomainfrom
nvp/agent-skills

Conversation

@NullVoxPopuli
Copy link
Contributor

@NullVoxPopuli NullVoxPopuli commented Jan 26, 2026

Propose adopting agent skills

Rendered

Summary

This pull request is proposing a new RFC.

To succeed, it will need to pass into the Exploring Stage, followed by the Accepted Stage.

A Proposed or Exploring RFC may also move to the Closed Stage if it is withdrawn by the author or if it is rejected by the Ember team. This requires an "FCP to Close" period.

An FCP is required before merging this PR to advance to Accepted.

Upon merging this PR, automation will open a draft PR for this RFC to move to the Ready for Released Stage.

Exploring Stage Description

This stage is entered when the Ember team believes the concept described in the RFC should be pursued, but the RFC may still need some more work, discussion, answers to open questions, and/or a champion before it can move to the next stage.

An RFC is moved into Exploring with consensus of the relevant teams. The relevant team expects to spend time helping to refine the proposal. The RFC remains a PR and will have an Exploring label applied.

An Exploring RFC that is successfully completed can move to Accepted with an FCP is required as in the existing process. It may also be moved to Closed with an FCP.

Accepted Stage Description

To move into the "accepted stage" the RFC must have complete prose and have successfully passed through an "FCP to Accept" period in which the community has weighed in and consensus has been achieved on the direction. The relevant teams believe that the proposal is well-specified and ready for implementation. The RFC has a champion within one of the relevant teams.

If there are unanswered questions, we have outlined them and expect that they will be answered before Ready for Release.

When the RFC is accepted, the PR will be merged, and automation will open a new PR to move the RFC to the Ready for Release stage. That PR should be used to track implementation progress and gain consensus to move to the next stage.

Checklist to move to Exploring

  • The team believes the concepts described in the RFC should be pursued.
  • The label S-Proposed is removed from the PR and the label S-Exploring is added.
  • The Ember team is willing to work on the proposal to get it to Accepted

Checklist to move to Accepted

  • This PR has had the Final Comment Period label has been added to start the FCP
  • The RFC is announced in #news-and-announcements in the Ember Discord.
  • The RFC has complete prose, is well-specified and ready for implementation.
    • All sections of the RFC are filled out.
    • Any unanswered questions are outlined and expected to be answered before Ready for Release.
    • "How we teach this?" is sufficiently filled out.
  • The RFC has a champion within one of the relevant teams.
  • The RFC has consensus after the FCP period.

@github-actions github-actions bot added the S-Proposed In the Proposed Stage label Jan 26, 2026
@johanrd
Copy link

johanrd commented Jan 27, 2026

Rendered link: https://github.com/emberjs/rfcs/blob/nvp/agent-skills/text/1165-ai-agent-skills.md

@mrloop
Copy link

mrloop commented Jan 27, 2026

The web-design-guidelines directs the agent to download the web interface guidelines from a URL, my initial thought was way not vendor https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md in this repo to isolate it, but maybe the benefits of up to date guideline out way any risk as the risk of anything malicious published to https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md is fairly unlikely?

@les2
Copy link
Contributor

les2 commented Jan 27, 2026

you could view this as a marketing / hype necessity -- frameworks need to advertise themselves as being the best for AI-driven development, right?

one "promise" of the LLMs is that they can understand information written for humans, which causes me to ask every time i write rules for these bots:
why not just document this stuff for humans (like in the readme) and let the AIs automatically discover and learn it?

as I read through your rules ... these are definitely useful for humans!!! they could be published as a website called "ember in action" or made into a book like in the old days.

another question is whether the need to write these rules -- as opposed to point to the documentation for the modules referenced -- reflects a problem with the existing API documentation, guides, and so on. basically, why isn't there an index with pointers to the canonical documentation, e.g.:

Use Ember Concurrency for User Input Concurrency
HIGH
Better control of user-initiated async operations
ember-concurrency, tasks, user-input, concurrency-patterns

Links:
- ember-concurrent doc link

finally, as a practical matter, all that matters is how well these rules work. i'm eager to try them!!!

@ef4
Copy link
Contributor

ef4 commented Jan 30, 2026

Similar to our discussion about the mcp server, versioning seems important. Maybe at the level of editions, or at the levels of releases, etc.

@NullVoxPopuli
Copy link
Contributor Author

NullVoxPopuli commented Feb 25, 2026

The current experience:

image

Would be good to instead do:

pnpm dlx skills add ember-tooling/agent-skills

I want this RFC to be more about the idea of adopting skills repos in general, not the actual contents. because the actual contents are going to be a living set of micro documents, and are 100% out of scope here.

I want everyone to feel like they can just contribute ideas they have, and not be gated by too much process -- since AI is wibbly wobbly, any potential improvement is a good one -- I don't thiiiiink this is something we can have CI test correctness / goodness of... so... it's literally just vibes. haha

@RobbieTheWagner
Copy link
Member

I think skills are very important right now, but they might be completely obsolete in the future, who knows. AI moves fast and I agree with @NullVoxPopuli that we should move fast and keep things fluid and easy to contribute to stay relevant.

@NullVoxPopuli
Copy link
Contributor Author

Similar to our discussion about the mcp server, versioning seems important. Maybe at the level of editions, or at the levels of releases, etc.

we can have LTS-based folders for subsequent versions

so when folks run the skills command, they'd be presented with:

❯ pnpm dlx skills add nullvoxpopuli/agent-skills
Packages: +1
+
Progress: resolved 1, reused 0, downloaded 1, added 1, done

███████╗██╗  ██╗██╗██╗     ██╗     ███████╗
██╔════╝██║ ██╔╝██║██║     ██║     ██╔════╝
███████╗█████╔╝ ██║██║     ██║     ███████╗
╚════██║██╔═██╗ ██║██║     ██║     ╚════██║
███████║██║  ██╗██║███████╗███████╗███████║
╚══════╝╚═╝  ╚═╝╚═╝╚══════╝╚══════╝╚══════╝

┌   skills
│
◇  Source: https://github.com/ember-tooling/agent-skills.git
│
◇  Repository cloned
│
◇  Found 2 skills
│
◆  Select skills to install (space to toggle)
│  ◻ ember-release
│  ◻ ember-lts-6.12
└

and we'll copy the ember-release files into new LTS folders as LTSes are declared

@RobbieTheWagner
Copy link
Member

Similar to our discussion about the mcp server, versioning seems important. Maybe at the level of editions, or at the levels of releases, etc.

we can have LTS-based folders for subsequent versions

so when folks run the skills command, they'd be presented with:

❯ pnpm dlx skills add nullvoxpopuli/agent-skills
Packages: +1
+
Progress: resolved 1, reused 0, downloaded 1, added 1, done

███████╗██╗  ██╗██╗██╗     ██╗     ███████╗
██╔════╝██║ ██╔╝██║██║     ██║     ██╔════╝
███████╗█████╔╝ ██║██║     ██║     ███████╗
╚════██║██╔═██╗ ██║██║     ██║     ╚════██║
███████║██║  ██╗██║███████╗███████╗███████║
╚══════╝╚═╝  ╚═╝╚═╝╚══════╝╚══════╝╚══════╝

┌   skills
│
◇  Source: https://github.com/ember-tooling/agent-skills.git
│
◇  Repository cloned
│
◇  Found 2 skills
│
◆  Select skills to install (space to toggle)
│  ◻ ember-release
│  ◻ ember-lts-6.12
└

and we'll copy the ember-release files into new LTS folders as LTSes are declared

Would we backport this to add stuff for like Ember 4 or only modern stuff?

@NullVoxPopuli
Copy link
Contributor Author

If there is community desire for that, it could be done, but it would make me want to reorganize the skills into categories, because much of what can be done today can be done in ember 4, and it's more feature based, such as template tag, and i wouldn't want to recommend hbs for anything if the user isn't using hbs

@MrChocolatine
Copy link
Contributor

Is Ember taking a direction into AI?

@NullVoxPopuli
Copy link
Contributor Author

ai or not, this project helps out humans as well, and is essentially partially superseding the cookbook RFC -- though it would be cool to render all these under https://cookbook.emberjs.com so we can complete that work

@cloke
Copy link

cloke commented Feb 25, 2026

100% support this! I can confirm that these skills and the other MCP rfc drastically improve the quality of agent output. As a non-scientific test I used a local qwen3-coder-next install "create me a modern button with ember". It did ok and I would say it was about 70% correct, but would have confused anyone not familiar with Ember. Then I added a layer that injected the skills and asked the same question and it was hard to find fault in the answer. As @RobbieTheWagner says AI moves fast and skills might not be a thing this time next year, but moving fast on these sorts of RFCs can only improve the ecosystem in the short term.

@NullVoxPopuli
Copy link
Contributor Author

tl;dr: for the agentskills standards that are forming https://github.com/mgechev/skills-best-practices (and the proposed repo mentioned in this RFC needs updating for maximum token efficiency)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

S-Proposed In the Proposed Stage

Projects

None yet

Development

Successfully merging this pull request may close these issues.

8 participants