Written by @inayayousfi, typed by GPT-5.6 Sol running in OpenCode.
Every call here is @inayayousfi's, and no agent acted on its own.
Problem
AGENTS.md repeats queue and subsystem-boundary guidance across its non-negotiable rules, preferred direction, avoid list, and queue architecture section.
The repeated wording makes the file harder to scan and creates several places that can drift apart.
Scope
- Keep one authoritative statement for each architecture constraint
- Separate mandatory rules from recommendations and examples
- Clarify when runtime events, commands, or direct calls are appropriate
- Preserve the modular composition, central state ownership, and Lua-readiness constraints
Acceptance criteria
- Each rule has one authoritative location
- Mandatory language is limited to actual non-negotiable constraints
- Event and command responsibilities are stated without repetition
- The file is shorter and easier to scan
- Application behavior remains unchanged
Written by @inayayousfi, typed by GPT-5.6 Sol running in OpenCode.
Every call here is @inayayousfi's, and no agent acted on its own.
Problem
AGENTS.mdrepeats queue and subsystem-boundary guidance across its non-negotiable rules, preferred direction, avoid list, and queue architecture section.The repeated wording makes the file harder to scan and creates several places that can drift apart.
Scope
Acceptance criteria