chore: add info for working with intrinsics to AGENTS.md

[psschwei]

Misc PR

Type of PR

• Bug Fix
• New Feature
• Documentation
• Other

Description

• Link to Issue: Fixes

Adds information for agents when working with intrinsics in this repo

Testing

• Tests added to the respective file if code was changed
• New code has 100% coverage if code as added
• Ensure existing tests and github automation passes (a maintainer will kick off the github automation when the rest of the PR is populated)

[github-actions]

The PR description has been updated. Please fill out the template for your PR to be reviewed.

[psschwei]

cc @generative-computing/mellea-intrinsics

[jakelorocco]

I think this is a good thing to add. I have a few vibe-based thoughts. Feel free to disagree with them.

[jakelorocco]

I worry that this piece of things might be too generic to regular intrinsics and not utilizing them in Mellea. The following is just based off vibes:

1. I would maybe preface with the how to use in Mellea stuff (the project resources below but maybe with more details here).
2. I would phrase this current working with intrinsics section as something like "how to add new intrinsics".

Since we utilize the "granite-common" formatters for intrinsics, the agent shouldn't actually need to know how to modify the context / messages, the library should do that for us. It just needs to know whether to call the high-level / convenience wrappers or the regular usage.

[psschwei]

Those are fair points. Referring to the model cards was something the intrinsics team suggested, but I believe that was mostly around intended usage and examples (in a previous PR we had used them in a counterproductive way). Let me revise a bit

[planetf1]

Not tested, but very much support the principle of better agent instructions.

However I wonder if this would be better as a skill in .agents/skills since it's targeted at something quite specific. That would keep the context down for general agent interactions. We'd just put an entry into AGENTS.md with something like

## 13. Intrinsics
Working with intrinsics (LoRA adapters for RAG, safety, calibration)? Use `/intrinsics-guide`.

I think much of the rest guide is about process - and common to most tasks. Though there may be errors equally relevant to consider for a skill - it's just we only just started adding them ;-)

No objections if you want to merge - we can skillify after

[psschwei]

However I wonder if this would be better as a skill in .agents/skills since it's targeted at something quite specific.

The idea was to get this in as a quick tactical fix, as the lack of intrinsics info was causing real problems. Turns out it didn't get in quickly 😄

I'd still be inclined to merge this and then do a update/refactor of AGENTS.md more broadly (it's also probably due an update, I had one open in #693 but that also wasn't merged) to take in the latest changes.