buremba

6 ms·

31.
by buremba 5mo ago
Yeah I have logs in Sentry, which also uses Postgresql.
32.
by buremba 5mo ago
All you need is Postgres until you scale into TBs of data. We use Postgresql as a durable workflow engine, vector search, time-series data, BM25 search, OLTP/OLAP engine, and a queue. It's basically the only dependency we have for
33.
by buremba 6mo ago
I don't usually like my cursor to be hijacked but man, I like this one.
34.
by buremba 6mo ago
Yes that's it. I actually just ask codex/claude code to look up the session id when I want to resume sessions cross harness, it's just jsonl files locally so it can access the full conversation history when needed.
35.
by buremba 6mo ago
Since prompt caching won't work across different models, how is this approach better than dropping a PR for the other harnesses to review?
36.
by buremba 6mo ago
I'm working on multi-tenant version of OpenClaw for organizations that has shared memory layer. It includes an entity based agent context layer that can be used as OpenClaw plugin and a sandbox runtime layer which uses just-bash with p
37.
by buremba 6mo ago
It playbook is that a model is too dangerous until a competitor releases a competing model that beats yours.
38.
by buremba 6mo ago
You can ask your agent to verify or review code. Just because people wrote code by hand, it doesn't mean you should trust
39.
by buremba 6mo ago
I get why they block OpenClaw and it makes sense but I wonder if they can actually detect OpenClaw calling Claude Code CLI using something like acpx. It's simply identical to how people use Claude Code locally.
40.
by buremba 6mo ago
does it matter?
41.
by buremba 6mo ago
Compliance tech company who doesn't know about open-source. Interesting.
42.
by buremba 6mo ago
I extensively use uv (IMO better than venv) but still it's Python specific and not universal. npm is much worse and native binaries are almost impossible to deal with multiple versions. nix is specifically targeting this use-case and i
43.
by buremba 6mo ago
They're composable but computers are not. Two skills might depend on a different version of a dependency which is pretty hard to maintain and their needs to be a deterministic system (agents are not) to resolve the conflicts and make s
44.
by buremba 6mo ago
I vote for taste
45.
MCP Is Overengineered, Skills Are Too Primitive(lobu.ai)
3 points by buremba 6mo ago|6 comments
46.
by buremba 8mo ago
CLI enables the actions to be made on behalf of you, the external service is not aware whether it's you or AI making the calls. With MCP, Sentry knows it's AI making the call so can be smarter about the security. There is many MCP
47.
by buremba 8mo ago
There is nothing wrong with MCP, it's just that stdio MCP was overengineered. MCP's Streamable HTTP with OAuth discovery is the best way to ship AI integration with your product nowadays. CLIs require sandboxing, doesn't hand
48.
by buremba 8mo ago
> but it only works if the proxy itself is hardened against prompt injection. Yes, I'm experimenting using a small model like Haiku to double check if the request looks good. It adds quite a bit of latency but it might be the right
49.
by buremba 8mo ago
AFAIK Claude Code doesn't inject all the MCP output into the context. It limits 25k tokens and uses bash pipe operators to read the full output. That's at least what I see in the latest version.
50.
by buremba 8mo ago
> Right now there's no way to have fine-grained draft/read only perms on most email providers or email clients. If it can read your email it can send email. > harder than you might think. openclaw found my browser cookies. (
51.
by buremba 8mo ago
My take is that agents should only take actions that you can recover from by default. You can gradually give it more permission and build guardrails such as extra LLM auditing, time boxed whitelisted domains etc. That's what I'm e
52.
by buremba 8mo ago
I think they should be aware that CC is big enough codebase that they can't vibe code anymore.
53.
by buremba 8mo ago
Sandboxes are not enough but you can have more observability into what the agent is doing, only give it access to read-only data and let it take irreversible actions that you can recover from. Here are some tips from building sandboxed mult
54.
by buremba 8mo ago
Page returns 404. ACP is great, indeed better to give pi-mono ACP than claude or codex directly. https://x.com/bu7emba/status/2026364497527513440
55.
by buremba 8mo ago
Nobody knows, including Anthropic itself I suppose
56.
by buremba 8mo ago
You technically still use CC, it's not via SDK but via CLI programmatically triggered via pi.
57.
by buremba 8mo ago
Technically you're still using claude CLI with this pattern so it's not 3P app calling Anthropic APIs via your OAuth token. Even if you would use Claude Code SDK, your app is 3P so it's in a gray area. Anthropic docs is inten
58.
by buremba 8mo ago
I spent 3 months adopting Codex and Claude Code SDKs only to realize they're just vendor lock-in and brittle. They're intended to be used as CLI so it's not programmable enough as a library. After digging into OpenClaw codeba
59.
by buremba 8mo ago
I’m working on the documentation right now but I had to build 3 prototypes to get here. :) After seeing Deno and Fly, I rewrote the proxy being inspired by them. I integrates nicely with existing MCP proxy so agent doesn’t see any MCP secre
60.
by buremba 8mo ago
Neat! I wasn’t aware that Docker has an embedded microVM option. I use Kata Containers on Kubernetes (Firecrackers) and restrict network access with a proxy that supports you to block/allow domain access. Also swap secrets at runtime s

More ›