Planning and specifications
What is a product requirements document?
Define product intent, review it with people and agents, and stop before technical design.
Approve an AI coding plan before execution
Choose the right planning depth, annotate the proposal, and inspect revisions before implementation.
Plan Diff: see what changed
Compare revised plans without rereading every unchanged section.
HTML vs Markdown for agent work
Choose durable text or an interactive explainer based on the decision readers need to make.
Living documentation for AI agents
Keep current system truth separate from the plans and decisions that produced it.
Agent instructions
Write a good AGENTS.md
Give agents exact commands, boundaries, and references without creating an instruction dump.
Compound plan feedback
Find repeated local feedback and turn the useful patterns into maintained instructions.
Design a coding-agent handoff
Connect returned work to the request, revision, evidence, risks, and next action.
Markdown and document collaboration
Annotate a website or HTML file
Import a URL or render local HTML for agent feedback, or use App Notes to annotate a live page.
Understand AI-generated code before shipping
Connect intended behavior, implementation, validation, and current documentation.
Code review
Local diff review for coding agents
Return line-specific feedback from a local browser diff.
What is a revision diff?
Understand cumulative review across pushes, rebases, and force-pushes.
What are code review artifacts?
Use screenshots, video, HTML, Markdown, and other evidence beside code.
Review pull request artifacts
Check whether visual and operational evidence matches the current revision.
Why AI pull requests are harder to review
Control review load without trusting a generated summary as proof.
AI-development workflows
Stay in the loop and own quality
Keep human attention at the points where intent becomes code and code becomes a shipping decision.
What is an AI artifact?
Distinguish inspectable outputs from chat responses and source records.
What is context rot?
Recognize when more context reduces an agent’s ability to use the right evidence.
What is an LLM codebase wiki?
Use generated repository maps without confusing them with maintained authority.
Long-horizon tasks for AI agents
Preserve state, evidence, and stop conditions across long-running work.
Run an AI agent night shift
Prepare a bounded goal and inspect real outputs the next morning.