Hacker News: jlaneve

New comment by jlaneve in "Claude Code 500s"

jlaneve — Tue, 17 Mar 2026 19:47:53 +0000

Just for Opus 4.6. I've switched back to Sonnet and it's working.

Data engineering skills and agent tools

jlaneve — Mon, 02 Feb 2026 18:52:14 +0000

Article URL: https://github.com/astronomer/agents

Comments URL: https://news.ycombinator.com/item?id=46859683

Points: 1

# Comments: 0

Molthub

jlaneve — Mon, 02 Feb 2026 03:51:37 +0000

Article URL: https://moithub.com/

Comments URL: https://news.ycombinator.com/item?id=46852205

Points: 3

# Comments: 1

Writing Airflow Dags with Excel and Minecraft

jlaneve — Thu, 20 Nov 2025 20:16:28 +0000

Article URL: https://www.astronomer.io/blog/abstraction-with-dag-factory-from-excel-to-minecraft/

Comments URL: https://news.ycombinator.com/item?id=45997104

Points: 2

# Comments: 0

AI Agents, Productivity, and Higher-Order Thinking

jlaneve — Thu, 13 Nov 2025 17:22:21 +0000

Article URL: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5713646

Comments URL: https://news.ycombinator.com/item?id=45917622

Points: 2

# Comments: 1

YC Arena

jlaneve — Tue, 30 Sep 2025 22:51:26 +0000

Article URL: https://www.ycarena.com/

Comments URL: https://news.ycombinator.com/item?id=45432221

Points: 1

# Comments: 0

DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Gen

jlaneve — Tue, 08 Jul 2025 13:12:43 +0000

Article URL: https://arxiv.org/abs/2506.20639

Comments URL: https://news.ycombinator.com/item?id=44499642

Points: 7

# Comments: 0

New comment by jlaneve in "What to build instead of AI agents"

jlaneve — Thu, 03 Jul 2025 01:28:46 +0000

We’re seeing this all the time - taking traditional workflow orchestration tools and instrumenting LLMs as part of it. It becomes a lot easier to build these because the complexity comes from a) the model, which frontier labs are making easy and b) productionizing a workflow, which workflow orchestration tools make easy. It’s also easy to recognize value because these workflows are often grounded in existing work and thus easy to measure.

We see these patterns do much so that we packaged it up for Airflow (one of the most popular workflow tools)!

https://github.com/astronomer/airflow-ai-sdk

Slack status: trouble uploading images and files

jlaneve — Wed, 02 Jul 2025 19:22:55 +0000

Article URL: https://slack-status.com/2025-07/53ea713200a3d4e6

Comments URL: https://news.ycombinator.com/item?id=44447789

Points: 1

# Comments: 0

New comment by jlaneve in "LLM Memory"

jlaneve — Wed, 02 Jul 2025 12:29:27 +0000

The link should be https://saxenauts.io/blog/persona-graph

Open-sourcing circuit tracing tools

jlaneve — Thu, 29 May 2025 17:16:54 +0000

Article URL: https://www.anthropic.com/research/open-source-circuit-tracing

Comments URL: https://news.ycombinator.com/item?id=44128101

Points: 161

# Comments: 20

Anthropic's circuit tracer is now open source

jlaneve — Thu, 29 May 2025 17:16:03 +0000

Article URL: https://github.com/safety-research/circuit-tracer

Comments URL: https://news.ycombinator.com/item?id=44128093

Points: 3

# Comments: 0

New comment by jlaneve in "Ask HN: What are you working on? (May 2025)"

jlaneve — Sun, 25 May 2025 21:28:46 +0000

On the home page right now it links to the Slash funding announcement 4 days ago, but the description looks way off

New comment by jlaneve in "N8n – Flexible AI workflow automation for technical teams"

jlaneve — Sat, 03 May 2025 23:50:44 +0000

Where do you think the gap between Airflow and what you need is?

Apache Airflow 3: the most significant release in Airflow's history

jlaneve — Tue, 22 Apr 2025 23:08:22 +0000

Article URL: https://www.astronomer.io/airflow/3-0/intro/

Comments URL: https://news.ycombinator.com/item?id=43767118

Points: 3

# Comments: 0

New comment by jlaneve in "12-factor Agents: Patterns of reliable LLM applications"

jlaneve — Wed, 16 Apr 2025 18:18:27 +0000

I've been most impressed with Pydantic AI [1], so much so that we ended up building an SDK around it specifically for LLM workflows on Airflow [2].

[1] https://ai.pydantic.dev

[2] https://github.com/astronomer/airflow-ai-sdk

New comment by jlaneve in "LLM Workflows then Agents: Getting Started with Apache Airflow"

jlaneve — Tue, 01 Apr 2025 15:45:27 +0000

Disclaimer: author of the SDK here.

Airflow actually uses decorators to indicate something is an explicit task in a data pipeline vs just a utility function, so this follows that pattern!

It also uses an "operator" under the hood (Airflow's term for a pre-built, parameterized task) which can be subclassed and customized if you want to do any customization.

New comment by jlaneve in "LLM Workflows then Agents: Getting Started with Apache Airflow"

jlaneve — Tue, 01 Apr 2025 15:44:07 +0000

Disclaimer: author of the SDK here.

It is _potentially_ more restrictive than writing pure Python functions, but the plus side is that we can interject certain Airflow-specific features into how the agent runs. And this isn't mean for someone who knows agents inside & out / wants the low-level customizability.

The best example of this today is log groups: Airflow lets you log things out as part of a "group" which has some UI abstractions to make it easier. This SDK takes the raw agent tool calls and turns them each into a log group, so you can see a) at a high level what the agent is doing, and b) drill down into a specific tool call to understand what's happening within the tool call.

To your point about the `@task.llm_branch`, the SDK & Pydantic AI (which the SDK uses under the hood) will re-prompt the LLM up to a certain number of attempts if it receives output that isn't the name of a downstream task. So there shouldn't be much finickiness.

New comment by jlaneve in "AI agents: Less capability, more reliability, please"

jlaneve — Mon, 31 Mar 2025 18:16:46 +0000

I appreciate the distinction between agents and workflows - this seems to be commonly overlooked and in my opinion helps ground people in reliability vs capability. Today (and in the near future) there's not going to be "one agent to rule them all", so these LLM workflows don't need to be incredibly capable. They just need to do what they're intended to do _reliably_ and nothing more.

I've started taking a very data engineering-centric approach to the problem where you treat an LLM as an API call as you would any other tool in a pipeline, and it's crazy (or maybe not so crazy) what LLM workflows are capable of doing, all with increased reliability. So much so that I've tried to package my thoughts / opinions up into an AI SDK for Apache Airflow [1] (one of the more popular orchestration tools that data engineers use). This feels like the right approach and in our customer base / community, it also maps perfectly to the organizations that have been most successful. The number of times I've seen companies stand up an AI team without really understanding _what problem they want to solve_...

[1] https://github.com/astronomer/airflow-ai-sdk

Build LLM Workflows Before Agents

jlaneve — Fri, 28 Mar 2025 16:13:29 +0000

Article URL: https://www.astronomer.io/blog/workflows-then-agents/

Comments URL: https://news.ycombinator.com/item?id=43507162

Points: 4

# Comments: 0