📝 docs: Add Get Involved page for new contributors #1

Manually merged

jflory7 merged 1 commit from add/get-involved into main

2026-09-21 18:59:49 +00:00

Author	SHA1	Message	Date
Justin Wheeler	b56b3c8eee	📝 docs: Add Get Involved page for new contributors Create a landing page that describes the four main contribution areas for the AI/ML SIG: RPM packaging, the Agent Skills Library, the AI Developer Desktop (Project Resistor), and user documentation. Each area explains what contributors would do and how to get started, with links to the relevant repos and resources. The page is designed to both welcome packagers and non-packager contributors, since the SIG's big-tent approach includes skills development, desktop building, documentation, and community engagement alongside the core packaging work. Closes: ai-ml/tickets#29 Assisted-by: Claude Opus 4.6 (1M context) Signed-off-by: Justin Wheeler <jwheel@redhat.com> 🚧 Address @lbarczio feedback: Better definition of Agent Skills I integrated Laura's feedback and made other minor revisions to integrate the opening paragraph better. ref: #1 (comment) Signed-off-by: Justin Wheeler <jwheel@redhat.com> 🚧 Address @lbarczio feedback: Try out and call them Agent Skills This commit does two things. Primarily, it addresses feedback from @lbarczio in PR ai-ml/docs#1 to add a call-out for getting involved by trying out Agent Skills and using them in real workflows, so that we can test them, improve them, and share the improvements with others. I also figured since we are building on a common spec, i.e., Agent Skills, it makes sense to consistently call them this proper (English) noun with capital letters. I believe this better communicates that we are building on a common spec, instead of forging our own totally new way of building these. Signed-off-by: Justin Wheeler <jwheel@redhat.com> 🚧 Address @lbarczio feedback: REVIEWING.md, call them A.S., tweak intro This commit primarily addresses Laura's final feedback in ai-ml/docs#1 of adding two new ideas for getting started with contributing to the Fedora Agent Skills Library with reviewing contributions and submitting improvements or new Agent Skills. I also continued the work in the previous commit of consistently calling them "Agent Skills" instead of just skills, to make it more consistent and clear that we are following the common spec. I also tweaked the intro paragraph slightly to reorganize things so that similar sentences were grouped together. Signed-off-by: Justin Wheeler <jwheel@redhat.com> 🚧 Address @tflink feedback: packaging scope, wording, FAS group path @tflink reviewed the RPM packaging and cross-cutting sections of the Get Involved page in ai-ml/docs#1 and requested three changes. The SIG doesn't package drivers for Fedora, so I dropped that word from the scope statement in the RPM packaging section. Tim also flagged "What you would do" as presumptuous across all four contribution areas: these are options a contributor could pick up, not a checklist they would necessarily do. Changed to "What you could do" in RPM packaging, Agent Skills, AI Developer Desktop, and User Documentation. Finally, the RPM packaging get-started list ended with "join the `ai-packagers-sig` FAS group," which read as an open invitation to request write access directly. @tflink's packaging experience is that groups like this prefer PRs and Matrix discussion first, with write access following from that track record rather than being requested up front. Reworded the bullet to reflect that path. ref: #1 (comment) ref: #1 (comment) ref: #1 (comment) ref: #1 (comment) ref: #1 (comment) ref: #1 (comment) Assisted-by: Claude Sonnet 5 (1M context) Signed-off-by: Justin Wheeler <jwheel@redhat.com> 🚧 Address @tflink feedback: note overlapping SIG membership @tflink pointed out in ai-ml/docs#1 that the PyTorch, Heterogeneous Computing, and ROCm Packaging SIGs currently share a lot of the same people. Rather than have a contributor try to figure out which of those SIGs actually owns a given piece of work, it's more useful for them to just start with the AI/ML SIG. Added two sentences to the note reflecting that. ref: #1 (comment) Assisted-by: Claude Sonnet 5 (1M context) Signed-off-by: Justin Wheeler <jwheel@redhat.com>	2026-09-21 14:55:26 -04:00

Author

SHA1

Message

Date

Justin Wheeler

b56b3c8eee

📝 docs: Add Get Involved page for new contributors

Create a landing page that describes the four main contribution areas
for the AI/ML SIG: RPM packaging, the Agent Skills Library, the AI
Developer Desktop (Project Resistor), and user documentation. Each area
explains what contributors would do and how to get started, with links
to the relevant repos and resources.

The page is designed to both welcome packagers and non-packager
contributors, since the SIG's big-tent approach includes skills
development, desktop building, documentation, and community engagement
alongside the core packaging work.

Closes: ai-ml/tickets#29

Assisted-by: Claude Opus 4.6 (1M context)
Signed-off-by: Justin Wheeler <jwheel@redhat.com>

🚧 Address @lbarczio feedback: Better definition of Agent Skills

I integrated Laura's feedback and made other minor revisions to
integrate the opening paragraph better.

ref: #1 (comment)

Signed-off-by: Justin Wheeler <jwheel@redhat.com>

🚧 Address @lbarczio feedback: Try out and call them Agent Skills

This commit does two things. Primarily, it addresses feedback from
@lbarczio in PR ai-ml/docs#1 to add a call-out for getting involved by
trying out Agent Skills and using them in real workflows, so that we can
test them, improve them, and share the improvements with others.

I also figured since we are building on a common spec, i.e., Agent
Skills, it makes sense to consistently call them this proper (English)
noun with capital letters. I believe this better communicates that we
are building on a common spec, instead of forging our own totally new
way of building these.

Signed-off-by: Justin Wheeler <jwheel@redhat.com>

🚧 Address @lbarczio feedback: REVIEWING.md, call them A.S., tweak intro

This commit primarily addresses Laura's final feedback in ai-ml/docs#1
of adding two new ideas for getting started with contributing to the
Fedora Agent Skills Library with reviewing contributions and submitting
improvements or new Agent Skills.

I also continued the work in the previous commit of consistently calling
them "Agent Skills" instead of just skills, to make it more consistent
and clear that we are following the common spec.

I also tweaked the intro paragraph slightly to reorganize things so that
similar sentences were grouped together.

Signed-off-by: Justin Wheeler <jwheel@redhat.com>

🚧 Address @tflink feedback: packaging scope, wording, FAS group path

@tflink reviewed the RPM packaging and cross-cutting sections of the Get
Involved page in ai-ml/docs#1 and requested three changes.

The SIG doesn't package drivers for Fedora, so I dropped that word from
the scope statement in the RPM packaging section. Tim also flagged "What
you would do" as presumptuous across all four contribution areas: these
are options a contributor could pick up, not a checklist they would
necessarily do. Changed to "What you could do" in RPM packaging, Agent
Skills, AI Developer Desktop, and User Documentation.

Finally, the RPM packaging get-started list ended with "join the
`ai-packagers-sig` FAS group," which read as an open invitation to
request write access directly. @tflink's packaging experience is that
groups like this prefer PRs and Matrix discussion first, with write
access following from that track record rather than being requested up
front. Reworded the bullet to reflect that path.

ref: #1 (comment)
ref: #1 (comment)
ref: #1 (comment)
ref: #1 (comment)
ref: #1 (comment)
ref: #1 (comment)

Assisted-by: Claude Sonnet 5 (1M context)
Signed-off-by: Justin Wheeler <jwheel@redhat.com>

🚧 Address @tflink feedback: note overlapping SIG membership

@tflink pointed out in ai-ml/docs#1 that the PyTorch, Heterogeneous
Computing, and ROCm Packaging SIGs currently share a lot of the same
people. Rather than have a contributor try to figure out which of those
SIGs actually owns a given piece of work, it's more useful for them to
just start with the AI/ML SIG. Added two sentences to the note
reflecting that.

ref: #1 (comment)

Assisted-by: Claude Sonnet 5 (1M context)
Signed-off-by: Justin Wheeler <jwheel@redhat.com>

2026-09-21 14:55:26 -04:00