frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: AI agent that runs real browser workflows

https://ghostd.io
2•heavymemory•1h ago
I’ve been experimenting with letting an AI agent execute full workflows in a browser.

In this demo I gave it my CV and asked it to find matching jobs. It scans my inbox, opens the listings, extracts the details and builds a Google Sheet automatically.

Comments

abraxas•1h ago
I was looking for a similar produc/project the other day. Alas my need is a Linux native version. You may want to consider it as Mac seems to be overserved by the agent harness supply while Linux is the opposite
heavymemory•1h ago
linux and windows support is on the way, i’ve designed it in a decoupled way, so should be straight forward.

Just need to see if people find this version useful

hkonte•48m ago
The CV-to-job-search demo is a good showcase. For multi-step workflows like this, one issue I've run into is that the agent instructions degrade across steps. The initial task description is clear, but by step 5 the model is interpolating intent from earlier context rather than following explicit instructions.

Structuring the task prompt into named blocks (objective, constraints, expected output format per step) before the workflow starts makes each step much more reliable. The agent has less to infer.

Built github.com/Nyrok/flompt to help with this, a visual builder that decomposes instructions into semantic blocks and compiles to Claude-optimized XML. Useful for defining the "task shape" before handing it to an agent.

heavymemory•31m ago
Yeah, instruction drift is a real problem in long agent chains. In this case the workflow gets decomposed into steps up front and each step is executed by a separate sub-agent.

So the model isn’t carrying the whole instruction chain across multiple steps, it’s just solving the current task. Similar pattern to what tools like Codex CLI or Claude Code do.

React library to add mentions in textarea

https://github.com/getnao/prompt-mentions
1•blef•54s ago•0 comments

How Do You Choose a Managed Security Service Provider (MSSP)?

https://msspproviders.io/
1•datacorp•1m ago•1 comments

XAI: Explainable Artificial Intelligence

https://www.darpa.mil/research/programs/explainable-artificial-intelligence
1•andsoitis•1m ago•0 comments

Show HN: ColumnLens – Query millions of rows in milliseconds on your Mac

1•hilti•3m ago•0 comments

The Found Report: A Look Inside America's Lost Luggage

https://www.unclaimedbaggage.com/pages/2026-found-report
1•SeanLang•4m ago•0 comments

Load-Bearing Walls

https://brianschrader.com/archive/load-bearing-walls/
1•sonicrocketman•5m ago•0 comments

Maybe All You Need Is the Friends You Made Along the Way

https://oooooooo.my/ayn.html
1•hpwfranks•5m ago•1 comments

Formalizing Data Structures and Algorithms with Agents

https://risemsr.github.io/blog/2026-03-06-autoclrs/
1•matt_d•5m ago•0 comments

Civic.games – Browser playable governance simulations

https://abeer1904.github.io/ai-race/
1•jimsojim•5m ago•0 comments

Ask HN: What Happened to xAI?

1•zof3•6m ago•0 comments

An OpenClaw skill for think-tank style analysis of crises like the Iran war

https://github.com/vassiliylakhonin/global-think-tank-analyst
1•vassilbek•7m ago•1 comments

Longest snake ever measured is over 23.5 feet long

https://www.popsci.com/environment/longest-snake-in-the-world/
1•bookofjoe•7m ago•0 comments

Six Selfish Reasons to Have Kids

https://kevinkelly.substack.com/p/six-selfish-reasons-to-have-kids
1•surprisetalk•10m ago•0 comments

The MIT Flea Swapfest 2026 Dates

https://web.mit.edu/w1mx/www/swapfest/flyer-2026.txt
1•oldnetguy•10m ago•0 comments

Ask HN: Optimizing Claude Code Workflow: Subscription or API Billing?

1•robeym•10m ago•0 comments

N times faster than C, Arm edition (2023)

https://blog.xoria.org/n-times-faster/
1•surprisetalk•10m ago•0 comments

Trapped in MS Office

https://ia.net/topics/trapped-in-ms-office
1•surprisetalk•10m ago•0 comments

Why haven't π±e, 2^e, π^e, or π^(√2) been shown to be irrational?

https://old.reddit.com/r/math/comments/d6xvk/why_havent_%CF%80e_2e_%CF%80e_or_%CF%802_been_shown_...
1•surprisetalk•10m ago•0 comments

We Tried Sitting on the Floor for 5 Years and Here's What Happened [video]

https://www.youtube.com/watch?v=L_9oU88UH_I
1•airstrike•11m ago•0 comments

Show HN: Gui.new – The Visual Layer for AI

https://gui.new
3•mynameisntbob•12m ago•1 comments

Ask HN: What is your current Agentic and/or Vibe coding setup?

1•sschueller•12m ago•0 comments

Black vs. White Gloss Chest: Pick a Tone to Match Your Style

https://dreamhomestore.co.uk/collections/chest-of-drawers
1•kevinwalton•13m ago•1 comments

Click-to-Release Strategy to Control Bond Cleavage and Prodrug Activation

https://onlinelibrary.wiley.com/doi/10.1002/anie.202515594
1•PaulHoule•14m ago•0 comments

Forge – OpenClaw for Enterprise

https://github.com/initializ/forge
1•ibreakthecloud•14m ago•0 comments

Sutton and Barto, Ch. 11: Off-Policy Methods with Approximation (Personal Notes)

https://chizkidd.github.io//2026/03/09/rl-sutton-barto-notes-ch011/
1•ibobev•16m ago•0 comments

Rebasing in Magit

https://entropicthoughts.com/rebasing-in-magit
8•ibobev•18m ago•0 comments

Vibe Coding: Build Fast, Think Slow

https://www.ssp.sh/brain/vibe-coding/
2•articsputnik•20m ago•0 comments

BullshitBench: Models Answering Nonsense Questions

https://petergpt.github.io/bullshit-benchmark/viewer/index.html
1•cebert•21m ago•0 comments

What happens when AI mediates everything?

1•spamsch•21m ago•2 comments

Stop Prompting. Start Approving

https://www.sustn.app/
3•mooreds•23m ago•0 comments