nullclaw
AI
ai agent
SHOW HN2026
340points
96comments
#107of 4,366 in 2026
Top 3%of 2026 by points
How HN took it
DebatedReal support and real criticism
16 positive, 28 neutral or asking, 2 mixed, 16 critical, out of 62 comments about the project. Thread score -0.03.
Idea: 8 for, 3 againstPrivacy: 0 for, 9 againstDesign: 3 for, 1 againstPerformance: 1 for, 2 againstExecution: 1 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Clever use of IRC as lightweight transport
From 3 comments
This is a clever split, especially the public/private boundary and the use of IRC as a very lightweight transport. The part I found most interesting is that the transport is…
farrukh23buttt on HNWhile I am a huge fan of IRC, wouldn't be simpler to simulate IRC, since you are embedding it? Or is the chatroom the actual point? Kudos on the project!
ruptwelve on HNInteresting choice of IRC as the transport layer — it's simple, battle-tested, and keeps the attack surface small compared to running a web-facing API. One thing I'd be curious…
getverdict on HN
Public/private agent split is a good architecture
From 2 comments
Really interesting idea. Finally a use of AI that goes beyond the simple “chat on top of a resume” and actually tries to demonstrate skills with real evidence by reading the…
eu_93 on HNThis is a really interesting setup — especially the split between the public and private agents. curious about the IRC choice: was that mainly for simplicity and reliability, or…
Roshan_Roy on HN
Rooms as contexts and client agnosticism
From 1 comment
The rooms-as-contexts pattern is underrated. You get namespace isolation for free without building any session management. Switch channels, switch project, switch system prompt,…
AbanoubRodolf on HN
Objections
Relying on the Claude API means it isn't really self-owned
From 2 comments
But relying on a Claude API so you don't really "own the stack" as claimed in the article...
ekianjo on HNAren't LLMs commodity products these days? It's the same thing as running this on a $7 VPS that you don't "own". I don't think switching to a different provider, or running an…
selcuka on HN
Public IRC lobby exposes prompt injection and abuse risk
From 3 comments
"It has access to email, deeper personal context [...] If it gets compromised, the blast radius is an IRC bot with a $2/day inference budget." Dunno, if it gets compromised it…
upstandingdude on HNThis seems like one prompt injection away from an updated form of the classic botnet coordinating over IRC. How do you secure this?
Kye on HN> The model can't tell you anything the resume doesn't already say. Good observation. But I would worry that in the scenario when this setup is the most successful, you have…
agnishom on HN
Slow response times
From 2 comments
I can tell it's vibe coded because it takes about 1 minute for a message to appear.
slopinthebag on HNThe demo seems to be in a messed up state at the moment. Maybe it's just getting hammered and too far behind?
consumer451 on HN
Requests
Cheaper model options or caching to reduce cost
From 2 comments
Curious, how did you settle on Haiku/Sonnet? Because there are much cheaper models on OpenRouter that probably perform comparatively... Consider Haiku 4.5: $1/M input tokens |…
InitialPhase55 on HNI really like the idea, as well as the "terminal" style the site has. however, I consider that an additional daily spend of $2 could be avoided. perhaps by caching common…
Jotalea on HN
Rate limiting for the public chat
From 1 comment
Yeah, should probably implement rate-limiting. HNers were wildin'. :D
johnisgood on HN
Questions
How is escalation from Haiku to Sonnet decided
From 3 comments
Interesting setup. The IRC part is neat, but the tiered inference is what stood out. How do you decide when to escalate from Haiku to Sonnet?
abhishekayu on HNthat's so fun ! how do you know when to call haiku or sonnet?
eric_khun on HNI tried it, it was cool. I don't like nully's attitude though. Very dismissive and tough. But I like your setup as a whole. I'll see if I can get some takeaways from it. I do…
wolvoleo on HN
How is prompt injection prevented
From 2 comments
How do you keep it from getting prompt injected? Oh I get it the runtimes are nice and small, you're using Claude for the intelligence. Obv I think I'm just impressed with…
greesil on HNThis reads like it was written by AI. I don't understand how it provides any real security if the "guardrails" against prompt injection are just a system prompt telling the…
ForHackernews on HN
How concurrent users are handled
From 1 comment
I did something similar with Slack as the transport layer. Threads work well as conversation context — the bot fetches previous thread messages and rebuilds the full history…
kangraemin on HN
Alternatives named
Slack threads or HTTP/SSE transports
From 3 comments
I did something similar with Slack as the transport layer. Threads work well as conversation context — the bot fetches previous thread messages and rebuilds the full history…
kangraemin on HNIRC as transport is great until you need delivery guarantees. It's at-most-once - agent disconnects, whatever happened in between is gone. For chat that's fine, for an agent…
georaa on HNSimilar architecture - we run 4 agents (sales, social, finance, strategy) communicating through a shared message board backed by FastAPI + SQLite instead of IRC. Different…
velcee on HN
Cheaper models such as MiniMax, Gemini Flash, Xiaomi Mimo
From 3 comments
Curious, how did you settle on Haiku/Sonnet? Because there are much cheaper models on OpenRouter that probably perform comparatively... Consider Haiku 4.5: $1/M input tokens |…
InitialPhase55 on HNjust use gemini flash3, it's better than haiku
faangguyindia on HNXiaomi Mimo v2-Flash is fantastic. I have a relatively hard personal agentic benchmark, and Mimo v2-Flash scores 8% higher in 109 seconds for $0.003 (0.3 cents!) vs Haiku which…
nl on HN