aiboost

Exactly, a reproducible research workbench should itself be reproducible.

Everything here is open source and MIT license. The model may still be a black box, of course you can use an open model if you prefer.

Solid Go codebase, especially the lock-free CAS rate limiting and explicit crypto.ZeroKey memory wiping. Curious how you handle token billing if an upstream SSE stream drops before the final usage chunk arrives?

I currently use the built-in caffeinate command in the terminal for this exact use case (keeping AI agents or large downloads running while away).

However, caffeinate fails when I physically close the MacBook lid (unless connected to an external display/power). Looking at your architecture, it seems Lockpaw uses IOPMAssertion from IOKit, which is the same API caffeinate uses.

Does Lockpaw have any workarounds for the lid-closed sleep behavior, or is the primary advantage here combining standard sleep assertions with visual shielding/input blocking? (Either way, solving the 'screen is on and fully accessible' problem of caffeinate is a huge win! Great app.)

Text files + Git > Vector DBs. Nice work. I'm curious about how this scales. As the notes/ directory grows over weeks or months, reading past daily logs will eat up the context window.

Thanks for checking the list — I’m maintaining this repo and will update it based on feedback.

Quick notes:

1. This is a curated index, not runnable code — entries have one-line descriptions and tags to help scanning.

2. If something’s missing or outdated, please open an issue or submit a PR with a 1–2 line description + link — I’ll review and credit contributors quickly.

3. I’m happy to add domain-specific starter paths if people suggest them here.

Thanks for any feedback or suggestions!

Hi, I'm the maintainer of Awesome-A2A.

Why: Google’s Agent2Agent protocol is still new and resources are scattered.

What's inside: 40+ SDKs across Go/Rust/Python/TS, demo apps, validation tools.

Gaps: Looking for production-scale examples and monitoring plugins — PRs very welcome.

Questions: What blocks you from trying A2A right now? Any real-world use cases you’d like us to cover?

Happy to answer anything about the curation process or A2A internals.