agentdev001

3 ms·

1.
by agentdev001 5d ago
Something like openshell is the answer here, to the point of GET being a write. The gap left here is what, imo, is something that MCP fits nicely- which is serving non-http resources with restrictions: databases for example.
2.
by agentdev001 5d ago
"without first having to solve the problem of effectively sandboxing Bash" Hopefully this is easier as time goes on. Of course- also policy on the egress
3.
by agentdev001 5d ago
Sounds ripe for vibe... sewing
4.
by agentdev001 5d ago
An example of this, Moonshot (kimi) open sourced this: https://kvcache-ai.github.io/AgentENV/latest/getting-started...
5.
by agentdev001 5d ago
"Is giving the agent a temporary scratchbox really that valuable?" Yes, but, wrong layer here. Giving the agent a computer use (a la bash) is what folks are after. A temporary sandbox with lots of control knobs and security bits i
6.
by agentdev001 6d ago
You must construct additional pylons
7.
by agentdev001 8d ago
> Vaguely the same result as RAG. Unless you’re in specific domains, you won’t beat handing the agent a shell and grep. This has been my conclusion as well, and I'm doing my best to try and back this up quantitatively. In a perfect
8.
by agentdev001 8d ago
Im not sure how this could make sense, considering the content of the paper.
9.
by agentdev001 8d ago
As far as I can tell, the paper says "bash capable", without ever describing what that means. How would one know whether a given model is "bash capable" or not? I would have to imagine, that Luna would very much fall int
10.
by agentdev001 9d ago
Some really broad assumptions here, and youre being unclear. "When you authorize OpenClaw to read your gMail, OpenClaw has the credential." I can only assume you are implying that the execution environment accessible by the model
11.
by agentdev001 9d ago
"The agent holds the credential" Huh? It shouldn't. Am I misunderstanding, or is this referencing poor practices? "Using the same tool every time prevents this" What does this mean? I looked at the project, im not s
12.
by agentdev001 9d ago
To say what the author said in another way- thats a garbage approach and completely misses the point. Sure, you can completely destroy the session by distilling it into a markdown- but thats like explaining what you did yesterday, vs having
13.
by agentdev001 9d ago
Yea- its just a matter of whether the work is delegated to the client or offered by the server. Making sure what is served is in a really quality schema is generally the most efficient path- ime
14.
by agentdev001 9d ago
> "I've found that it's so good that I cancelled building an an MCP server and any skill. Just a well documented GQL schema. It's pretty remarkable." This gave me a chuckle, there's some subtle irony here- e
15.
by agentdev001 11d ago
Absolutely love openshell as a solution, I really hope Kube support moves out of experimental some time in the near future. Cool solution here- to a problem that I imagine is probably impossible to get 100% Looking forward to 0.1.0 release
16.
by agentdev001 13d ago
IIRC, openai enabled the ability for subagents to use different models than the parent- maybe a month~ ago.
17.
by agentdev001 15d ago
And replying with some more thoughts after reading the comments here. To me, this feels similar to when models started passing the line of (imo) "good enough" to start building much more capable agents. The release of this (gpt-li
18.
by agentdev001 15d ago
Reposting my comment from https://news.ycombinator.com/item?id=49646963 Congrats on the launch here. I've been messing with this over the last few hours- super super cool. I was excitedly awaiting this hitting the API,
19.
by agentdev001 15d ago
Native integration with their SDK.
20.
by agentdev001 15d ago
Congrats on the launch here. I've been messing with this over the last few hours- super super cool. I was excitedly awaiting this hitting the API, because ofc there wasn't a super high fidelity option for drop-in voice interface i
21.
by agentdev001 16d ago
Your comment frustrates me! With that said, how about Discord? I have high speed internet 90% of my waking life, high enough for Discord to be high fidelity. I actively avoid using discord to join calls on my phone- because Bluetooth absolu
22.
by agentdev001 18d ago
Until either of these cases happen: A: Devs with an online presence stop using Anthropic models B: Anthropic catches up to OpenAI in terms of per-token efficiency, and average token total for final-output We will continue to see posts such
23.
by agentdev001 21d ago
Secrets shouldn't be plain text in project directory >:[
24.
by agentdev001 22d ago
I am not using MCP in be production- but, my team is. My team also produces MCP servers for other teams, and I find it a bit maddening. I wonder if anyone can relate to my experience here. It feels like there is a significant amount of bagg
25.
by agentdev001 22d ago
> "it makes things a tiny bit easier than interfacing directly via api" By what metric? I would expect that a thin API client (with readable code) is generally going to out-perform a tool-surface which you don't have the a
26.
by agentdev001 22d ago
Sounds like you might benefit from running a custom harness then, no? I can't imagine for a task such as that- that codex is the best option.
27.
by agentdev001 24d ago
3.8 Flash tomorrow apparently, WSJ says google insiders say on-par with Opus 5. Time will tell.
28.
by agentdev001 24d ago
This is moreso about the (human-intended) tools, data, and environments you have available to you. Wanna do defense? Get more telemetry. Wanna do red? Get solid test-bed environments. Mature infosec programs are benefiting the most, good-gu
29.
by agentdev001 24d ago
High, X-high, and Max are all on the $/intelligence Pareto
30.
by agentdev001 24d ago
Could be risky. Yet goal solution.

More ›