mehdibl

So to free up space you delete folders instead of moving your familyvphotos that you don't have backup.

Can we stop a bit this all evil Microsoft fault?

And the author have a solution. Yeah those headline are buzzing.

Docker is hard to setup. The author made a nice solution but not sure if he know devcontainer and what he can do. You do the setup once and you roll in most dev tools. I'm still surprised the effort people put in such solution ignore the dev's core requirements, like sharing the env they use in a simple way. You used it to have custom env and isolate the agent. You want to persist your credentials? Mount the target folder from home or sl into a sub folder. Might be knowledge. But for Linux or even Windows/Mac as long you don't need desktop fully. Devcontainer is simple. A standard that works. And it's very mature.

Isn't this mainly a skill injected in the context? Rather a model/platform specific feature?

Lovely you discovered devcontainers.

Ollama is quite a bad example here. Despite popular, it's a simple wrapper and more and more pushed by the app it wraps llama.cpp.

Don't understand here the parallel.

You can improve a lot the success rate with providing HELM and clear instructions with the tool description.

Over a year ago had a lot of issues and the description and example was the difference between 30-50% failure to 1%!

So I'm surprised a bit about the point. May be I'm missing it.

Why in nushell? Not in go?

But I like the idea and principle. OSS need this and it's traded very lightly.

Devcontainer is not docker and do the trick!

How to do agentic workflow like 2 years ago.

Claims as always misleading as they don't show the context length or prefill if you use a lot of context. As it will be fun waiting minutes for a reply.

Issue llama.cpp is shipped and no more updated. Or you need to download it all.

This is killing me with complexity. We had agents.md and were supposed to augment the context there. Now back to cursor rules and another md file to ingest.

May be Mozilla have a point.

Chat apps, are replacing a lot the old way we consume information and search. That is mostly made thru browser. So I see the vision is follow this transformation to keep market share and offer an alternative to big players.

Mozilla and Firefox loosing market share and revenue too and that could bite back.

Ok and then? Those models were not trained for this purpose.

It's like the last hype over using generative AI for trading.

You might use it for sentiment analysis, summarization and data pre-processing. But classic forecast models will outperform them if you feed them the right metrics.

Previous report blaming TPlink slow to patch a CVE were already outdated as the CVE got patched. Yes TPlink are recieving updates if the products are not EOL. And even US products when EOL are vulnerable.

Seem more heavy lobbying to get their US marketshare here rathar than looking for secure products.

Also the report from checkpoint over firmware used to attache EU, the malware is firmware agnostic. As it can be used for other hardware.

Before investing in instruments, you should have a solid static analysis, unit tests, integration test and so on. Logging help flagging issues post deployment but you can catch a lot if you test.

What matter is not context or the recod token/s you get.

But the quality for the model. And it seem Grok pushing the wrong metrics again, after launching fast.

We are still having to read this again in 2025? Some will never get it.

It's a Qwen 3 MoE fine tune...

The funny part the 5090 mobile is more like a 5080 desktop edition!

See the Alienware laptop flagged as 5090 while it's ‎"GeForce RTX 5090 24 GB GDDR7" as laptops can't sustain the TDP and RTX XX90 full power. For AI an external GPU is less costly option.

The gain is only in prefill and if the task/output is complex the gain will be totally minor. So the numbers are quitly exagerated here based on a prompt that is taking less than 2s to decode. So I guess we are not here doing complex tasks with 100's or 1000 token output. For the cost of an M3 Ultra + DGX the gain seem minimal and most of all, exo didn't clarify the model used here and it's for sure not a dense model or an MoE with 1B or 2B experts otherwise the mac ultra too will suffer a lot and the layers will be bigger!

Why only run inside docker? How about the dev's using devcontainer already!

Errors are normal and happen ofter. You need to focus on providing it ability to test the changes and fix errors.

If you expect one shot you will get a lot of bad surprises.

You can do copy and paste if you offer it a tool/MCP that do that. It's not complicated using either function extraction with AST as target or line numbers.

Also if you want it to pause asking questions, you need to offer that thru tools (example Manus do that) and I have an MCP that do that and surprisingly I got a lot of questions and if you prompt, it will do. But the push currently is for full automation and that's why it's not there. We are far better in supervised step by step mode. There is elicitation already in MCP, but having a tool asking questions require you have a UI that will allow to set the input back.

Ollama default to Q4 usually and 8/16k context and not the 1M context

Bad idea to use agents for prompts. I find them very very noisy and bloated.

This is a vendor selling a solution for "hypothecal" risk not seen in the WILD!

More seen as buzz article about how it could happen. This is very complicated to exploit vs classic supply chains and very narrow!

This is quite convoluted. Not seen in the wild and comments don't trigger prompt injection that easily.

Again you likely use vscode. Are you checking each extension you download? There is already a lot of reported attacks using vscode.

A lot of noise over MCP or tools hypothetical attacks. The attack surface is very narrow, vs what we already run before reaching Claude Code.

Yes Claude Code use curl and I find it quite annoying we can't shut the internal tools to replace them with MCP's that have filters, for better logging & ability to proxy/block action with more in depth analysis.

how are you going to get "adversarial attacks" with prompt injection. If you don't fetch data from external sources. Web scraping ( you can channel that thru Perplexity by the to sanitize it). PR reviews, would be fine if repo is private.

I feel this is overly exagerated here.

There is more issues that are currently getting leverage to hack with vscode extension than AI prompt injection, that require a VERY VERY complex chain of attack to get some leaks.

1. Author missed API existed. Easier to consume data/fetch report. You can still use LLM to process text report to extract data/sentiment indeed so the browser it self run already apps that heavily use API.

2. AI workflow, to fetch external data and allow AI to do it need tools, so it can dynamicly fetch or you will need to grab the data and inject it in the context.

3. MCP is a transport layer, you can have tools without MCP. You bake them in.

I feel there is some points that are mixed up.