rane

My similar workflow within Claude Code when it gets stuck is to have it consult Gemini. Works either through Gemini CLI or the API. Surprisingly powerful pattern because I've just found that Gemini is still ahead of Opus in architectural reasoning and figuring out difficult bugs. https://github.com/raine/consult-llm-mcp

Now what if it didn't pollute browser history

Depends on Anthropic has implemented it I guess. I haven't had it activate yet despite the prompt to install an LSP.

You don't want to run tests after every file change, because that will distract Claude from finishing whatever it's doing and add noise to the context window. Of course the tests will be broken if Claude hasn't finished the full change yet.

Running tests makes most sense on the Stop hook event, but personally I've found CLAUDE.md instruction of "Run `just check` after changes" to be effective enough. The Stop hook has the issue that it will run the checks every time Claude stops responding, even after without any changes.

Exactly. This is why the workflow of consulting Gemini/Codex for architecture and overall plan, and then have Claude implement the changes is so powerful.

It's not enough. Sometimes skills just randomly won't be invoked.

Awesome if that's the case

Have used Alfred for 10+ years at this point. Some colleagues are hyped about Raycast, but to me the pricing model is a joke. Pay (monthly) for AI - how about I bring my own API key? Pay (again, monthly) for unlimited clipboard history - lol. Free plan, "Free, forever". Yeah, until it isn't.

Alfred isn't the shiniest thing anymore but it's stood the time remarkably well, something I value very highly for tools as central to my workflow as Alfred.

tmux users might find this useful: https://github.com/raine/workmux

Said policy:

    * Any contribution of any LLM-generated content will be rejected and result in
      an immediate ban for the contributor, without recourse.

For good measure, I tried:

    description: CRITICAL: Use when writing bash scripts

Surprisingly no effect either. I would've thought adding "CRITICAL" would somehow promote that instruction in the sea of context.

So far not impressed with CC's ability to invoke skills automatically.

I made a skill with the unambiguous description: "Use when creating or editing bash scripts"

Yet, Claude does not invoke the skill when asked to write a bash script.

https://gist.github.com/raine/528f97375e125cf97a8f8b415bfd80...

I tried. Claude Code can't enable a disabled MCP on the fly.

In the AI image generation scene, is there anything solid yet in the way of generating vector illustrations for apps?

I tried in my relatively small project.

    ~/c/l/web % ck --sem 'error handling'
    ℹ Semantic search: top 10 results, threshold ≥0.6
    ⠹ Searching with semantic mode...

All I got was spinning M2 Mac fan after a minute, and gave up.

Cool. Some AI fluff can be detected in the README.

For example under the "Why CK?" section, "For teams" is of no substance compared to "For developers"

> People keep saying that Gemini 2.5 Pro can solve some problem that Sonnet 4 cannot

Most definitely can. It's insane how well just telling Claude to ask help from Gemini works in practice.

https://github.com/raine/consult-llm-mcp

Disclaimer: made it

Getting kicked out from open source community for lying about using AI? Haha, good one.

Why would these people disclose their use of AI? These are not responsible and thoughful users of AI. The slop producers won't disclose, and the responsible users who produce high quality PRs with AI will get the "AI slop" label. At this point, why even disclose if the AI-assisted high-quality PR is indistinguishable from having been manually written (which it should be)? No point.

Using AI to generate code in a PR does not necessarily mean however that the user has not taken time to understand the changes and is not willing to learn. There are AI users who generate whole files without understanding the contents, and then there are AI users who generate the exact same files but have known in advance what they want, and merely use AI as a tool to save typing.

The intention here seems to be to filter out low quality submissions for which the only purpose is to only pimp Github resume for having contributions in highly starred repo. Not sure if the people doing that will be disclosing use of AI anyway.

> I doubt a PR is going to be buried if it's useful, well designed, good code, etc, just because of this disclosure

If that were the case, why would this rule be necessary, if it indeed is the substance that matters? AI generated anything has a heavy slop stigma right now, even if the content is solid.

This would make for an interesting experiment to submit a PR that was absolute gold but with the disclaimer it was generated with help of ChatGPT. I would almost guarantee it would be received with skepticism and dismissals.

Using Anki myself for language learning and can definitely recognize the pitfall of unrelated cues triggering the recall, but so far it's been working quite well. I'm mainly using it to practice output though, so the front side of the card is in English.

On semi-related note, currently making on a language app (gengengo.com) if anyone wants to check it out.

How will they make money?

Is it though? Won't AI make the barrier of entry to indie game dev even lower, as assets and code will be able to be created effortlessly.

There's this one new cool technology that could be just the thing for converting the HTML to use web framework of preference.

Maybe our brains are wired different but reading and reviewing code is way faster for me than writing it.

You get much better results in CC as well if you're able to give the relevant files as a starting point. In that regard these two tools are not all that different.

I don't even write the outline myself. I tell CC to come up with a plan, and then we iterate on that together with CC and I might also give it to Gemini for review and tell CC to apply Gemini's suggestions.

Having used Claude Code extensively for the last few months, I still haven't reached this "until it isn't" point. Review the code that comes out. It goes a long way.

Aider will follow through a single, rather specific request, from the user, which will generally result in changes one or more files that you need to explicitly add to the context.

Claude Code on the other hand, based on user's request, is capable of:

- researching the codebase to understand what needs to be done

- planning a TODO list for itself

- implementing the TODO list item by item

- running tests or typecheck etc. and iterating until checks pass

Basically if you give a sufficiently detailed plan as markdown to CC, it is capable of autonomously implementing the plan from start to finish, and that could even take something like 20 minutes.