Fable is back, here's how I use it in Cursor
Fable as orchestrator with Composer workers, ultra long horizon cloud agents, and why rotating frontier models keeps your sense of what good looks like current
anyblockers
Observations from tinkering with AI
Latest posts
Human in the /loop
The setup I use to run long agent loops: a definition of done the agent can score, stop conditions, and pings so I know when to lean in
Don't let your agent guess, give it runtime context
How Debug Mode gives agents runtime evidence: instrument the code, reproduce the bug, read the logs, and fix the root cause instead of guessing
Optimizing your dev environment for coding agents
Primitives, guardrails, and enablers: how to prepare your codebase and environment so coding agents can do real work and verify it themselves
Your Mac mini is now a Cursor agent
Setting up a self-hosted Cursor worker on a Mac mini with access to iMessage, and letting an agent text my friends
Building CLIs for agents
Patterns that make CLIs work better for coding agents: non-interactive flags, useful --help with examples, idempotency, dry-runs, and predictable command structure
Towards self-learning coding agents
An experimental plugin that reads past chats, extracts learnings and preferences, and stores them in AGENTS.md so agents remember your feedback
Agentic codeowners
How we route PR reviews by risk with cloud agents: auto-approving low-risk changes and assigning the most relevant reviewers to high-risk ones
End of content perfectionism
Stop polishing everything to death. Be more human, embrace imperfection, and start livestreaming.
The problematic path of least resistance for coding agents
How AI coding agents often treat symptoms instead of root causes, leading to technical debt. Some preventative measures to ship better code.
Working with coding agents
What I have observed working well with AI coding agents during 2025 and practical tips to get the most out of them
GPT-5 Prompting Guide (with OpenAI)
A comprehensive guide to prompting GPT-5, covering best practices, techniques, and examples for getting the most out of the model.
Using query intent to boost retrieval results
Enhance your retrieval systems by determining user query intent and tailoring search strategies accordingly.
18 observations from using Cursor for 6+ months
My experience with Cursor over the last 6+ months on what works, what doesn't, and what I'm still figuring out
Building RAG with Postgres
A step by step guide to building a RAG system with Postgres
Postgres as a search engine
Build a retrieval system with semantic, full-text, and fuzzy search in Postgres to be used as a backbone in RAG pipelines.
How we systematically analyze user feedback with LLMs
A walkthrough of how we built an LLM-powered system to analyze user feedback at Qualtive. From planning to implementation, including our successes, failures, and unexpected challenges.
The idea of internal no-code platforms
What if we let LLMs orchestrate technical primitives & templates when building software?
Postgres is all you need, even for vectors
Exploring the versatility of Postgres for vector storage and querying with pgvector
Avoid the copilot pause
Tips to maximize coding flow and minimize distractions when working with coding copilots
How I became 3x more productive in 30 minutes with Cursor
How Cursor made me more productive and kept my flow state alive
Streaming JSON & text in one request
Learn how to stream JSON and text in one request using ReadableStream which will enhance your app's UX