These sources are about delegation in practice. They describe routines, permission limits and harnesses around coding agents. Many come from companies that sell these systems.
A Stupid Idea for AI Alignment We Came up with by Looking at the List of Specification Gaming Behaviours
A tongue-in-cheek essay proposing to align AI by making it want to die.
A spec the agent can't run isn't a spec yet
A short case for acceptance criteria that an agent can execute.
An Empirical Study: AI Agent Rules Need Context and Layered Enforcement
A blog post summarizing the ActPlane study of agent rules and kernel-level enforcement.
Show HN: Kastor – Terraform-style specs for AI agents
A Hacker News thread where a developer introduces Kastor, a spec language for agents.
Understanding Spec-Driven-Development: Kiro, spec-kit, and Tessl
An article defining spec-driven development and describing the author's trials of three tools.
I rewrote PostHog's SQL parser, 70x faster, while barely looking at the code
A Hacker News thread about a PostHog blog post on an AI-written, faster SQL parser.
Treat Agent Output Like Compiler Output
A company blog post arguing that agent-written code should be verified, not read.
Advanced Context Engineering for Coding Agents
A workflow that uses files to carry state between fresh agent contexts.
A Software Engineer's Guide to Agentic Software Development
A GitHub engineer describes a three-step routine for handing tech debt tasks to coding agents.
Agent Draw: An agent draws while you talk, built on TLDraw
A tutorial on building a tool that draws on a canvas from speech.
Agent Skills
A blog post explaining the design of the author's Agent Skills repo for coding agents.
Agentic Monitoring
A blog post arguing that AI agents are well suited to monitoring production systems.
Building Reliable Agentic AI Systems
A case study of a Bayer research assistant built with agents, retrieval and SQL generation.
Building an Advanced Agentic Harness
A tutorial that upgrades a basic agent loop into a planned, parallel, checked system.
Effective harnesses for long-running agents
An engineering post on helping coding agents make steady progress across many context windows.
Embracing the parallel coding agent lifestyle
A blog post on when running several coding agents at once is worth doing.
Harness engineering: leveraging Codex in an agent-first world
An OpenAI engineering post on building a product with Codex and no hand-written code.
How much can you delegate to agents?
A newsletter post offering four levels for deciding how much to delegate to agents.
Introducing Kitesurf: The agent-first browser that runs in V8 isolates on Cloudflare Workers
A Cloudflare post announcing a lightweight browser built for AI agents on its Workers platform.
My fireside chat about agentic engineering at the Pragmatic Summit
A weblog post of highlights from a fireside chat on agentic engineering.
Show HN: Juggler – an open-source GUI coding agent, by the creator of JUCE
A Hacker News launch thread for Juggler, a graphical coding agent with tree-shaped sessions.
Skill Issue: Harness Engineering for Coding Agents
A company blog post on configuring coding agents through skills, sub-agents, hooks and tests.
Technical Deep Dive on Streaming AI Agent Desktop Sandboxes: When Gaming Protocols Meet Multi-User Access
A company blog post on streaming agent desktops with a game streaming protocol.
The 8 Levels of Agentic Engineering
A blog post that sorts AI coding practice into eight levels.
The Coming Loop
A blog post on harness-level loops around coding agents and the unease they cause.
The Harness Is the Thing
A personal blog post claiming that the harness around the models matters most.
The quality of AI-assisted software depends on unit of work management
A blog post arguing that breaking AI coding work into small, human-legible units drives quality.
Unreal Agent
A company post announcing an agent harness that runs tool calls asynchronously.
Vibe Coding
A personal blog post on handing a fractal webpage to a coding agent.
Vibe engineering
A blog post proposing a name for professional, accountable work with coding agents.
agents need control flow, not more prompts
A short blog post arguing that reliable agents need control flow in code, not prompts.
Surprisingly Fast AI-Generated Kernels We Didn’t Mean to Publish (Yet)
A Stanford CRFM post reporting fast GPU kernels written by language models.
Logic Pro MCP Server for Claude, Cursor, and AI Agents
The README of an MIT-licensed server that lets AI agents control Logic Pro through MCP.