tiny-automates

5 ms·

1.
TinyFish Web Search and Fetch are now free, everywhere(tinyfish.ai)
5 points by tiny-automates 5mo ago|0 comments
2.
by tiny-automates 5mo ago
fetch : https://www.tinyfish.ai/blog/production-grade-web-fetching-f...
3.
Building a Browser for the Agent Era(tinyfish.ai)
6 points by tiny-automates 5mo ago|2 comments
4.
by tiny-automates 5mo ago
hey hn, we've been building the web infrastructure for ai agents at tinyfish. our agent primitive was already out. just went live with 3 new primitives - search, fetch, and browser - all under one api. the team wrote some details on h
5.
by tiny-automates 6mo ago
thanks! glad the visuals helped, we wanted the data to speak for itself.
6.
by tiny-automates 6mo ago
yeah that's basically the whole thesis compressed into six words. we kept trying to articulate why the CLI worked better and it always came back to that - MCP does discovery really well, no reason to throw that away. but the moment you
7.
by tiny-automates 7mo ago
the finding that buried everyone: moving "allow yourself to be shut down" from the user prompt to the system prompt made models sabotage more often, not less. Grok 4 went from 93% to 97%. system prompts are supposed to be the high
8.
Ask HN: How do you maximize your luck surface area?
2 points by tiny-automates 7mo ago|2 comments
9.
by tiny-automates 7mo ago
i'm building TinyFish ( https://tinyfish.ai )- browser automation APIs built for AI agents. most agent frameworks treat the browser like a human would (screenshot, click, pray), which breaks constantly. TinyFish provides stru
10.
by tiny-automates 7mo ago
agreed, and i'd go further - the harness is where evaluation actually happens, not in some separate benchmark suite. rhe model doesn't know if it succeeded at a web task. the harness has to verify DOM state, check that the right e
11.
by tiny-automates 7mo ago
the harness being "9 lines of code" is deceptive in the same way a web server is "just accept connections and serve files." the hard part isn't the loop itself — it's everything around failure recovery. when a
12.
Security audit of Browser Use: prompt injection, credential exfil, domain bypass(arxiv.org)
2 points by tiny-automates 7mo ago|1 comments
13.
by tiny-automates 7mo ago
the planner-executor isolation point is what stood out to me. right now most browser agent frameworks treat the LLM as both the decision-maker and the one processing untrusted content — so a prompt injection in page content can hijack the e
14.
Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs(arxiv.org)
544 points by tiny-automates 7mo ago|366 comments
15.
by tiny-automates 7mo ago
agree that this is a protocol-level issue, not framework-specific. but the "all external tool calls require confirmation prompts" mitigation doesn't really apply here - the exfil happens without any tool call. the model just
16.
by tiny-automates 7mo ago
the unfurling vector is elegant because it exploits a feature that predates LLMs entirely, link previews were designed for human-shared URLs where the sender is trusted. once an LLM is generating the message content, the trust model breaks
17.
by tiny-automates 7mo ago
that's the user-facing definition but the implementation distinction matters more. "takes longer than you're willing to wait" describes the UX, not the architecture. the engineering question is: does the system actually
18.
by tiny-automates 7mo ago
"background job" is actually the more honest framing. the interesting design question you're pointing at, what happens when it wants attention, is where the real complexity lives. in practice i've found three patterns:
19.
by tiny-automates 7mo ago
frontend QA is exactly where i've seen the biggest ROI with browser agents. the gap with Playwright MCP specifically is that it assumes the agent can reason about CSS selectors and DOM state, which breaks constantly on anything with dy
20.
by tiny-automates 7mo ago
the abstraction level argument is spot on. i've been working on browser automation for AI agents and the biggest lesson has been that exposing Playwright-level primitives to a foundation model is fundamentally the wrong interface. the
21.
by tiny-automates 7mo ago
i go back and forth on this. when i'm working on something where the hard part is the actual algorithm, say custom scheduling logic or a non-trivial state machine, i need my hands in the code because the implementation is the thinking.
22.
by tiny-automates 7mo ago
i've been building agent tooling for a while and this is the question i keep coming back to. the actual failure mode isn't messy code, agents produce reasonably clean, well-typed output these days. it's that the code confiden
23.
by tiny-automates 7mo ago
this is close to what i've landed on too. the pre-commit hook is non-negotiable. i've had Claude Code report "all checks pass" when there were 14 failing eslint rules. beyond the static analysis though, i keep hitting a