Recommended contribution strategy for new AI tool integrations

[waewoo]

Hi,

With recent integrations such as Kilo Code (#745, #992) and Antigravity (#957, #958, #960, #961, #962), we've seen different approaches to introducing new AI tool support into AIDD.

Both approaches have their advantages, and I think it would be useful to clarify the preferred contribution strategy for future integrations.

A few questions:

1. PR granularity: Should integrations be split by capability (CLI support, skills, rules, agents, hooks, memory, generators), or is a broader PR acceptable when these capabilities form a coherent feature?
2. Stacked vs. independent PRs: Are stacked PRs preferred for dependent capabilities, or should contributors aim for independently mergeable PRs?
3. Cross-tool improvements: Should shared infrastructure changes discovered during an integration be extracted into separate PRs?
4. Validation expectations: What level of unit, integration, and real-host runtime testing is expected before merging?

The goal isn't to revisit existing contributions, but to establish a consistent and contributor-friendly approach for future AI tool integrations.

Would it make sense to document these guidelines in CONTRIBUTING.md or the maintainer documentation?

Interested in hearing the maintainers' preferred approach.