429 is not an error, it is a schedule
Transient and terminal failures need opposite responses and look identical. How to tell them apart, and why partial progress must be persisted before it is believed.
All posts
A gate that fixes what it finds
An auto-healing loop must fix the source, never the symptom, and must never edit the thing that found the problem. Confidence floors and bounded rounds.
Six guardrails that stop my AI agent mid-command
Six PreToolUse hooks that refuse an agent's command before it runs. What each one caught, and why a guard the agent can talk its way past is not a guard.
Skills, tools, and the difference that matters
Progressive disclosure across four tiers: what must be loaded to decide, what loads to execute, and what should never load unless someone asks for it.
Stop parsing your AI's prose. Make it return a schema you can assert on.
Schema-forced returns move validation to the tool call, so the model retries instead of the caller parsing. What that buys, and the field that lied in the right format.
Subagents are context isolation, not parallelism
Every dispatched worker pays a fixed setup cost before it does anything. Where the crossover sits, and why a ten-item batch ran faster serially.
What to log when the actor is a model
Agent telemetry that survives the session: log the measurement rather than the verdict, and build a schema with no slot for the things that must not travel.
One tool call can destroy your agent's context window
What a tool returns is a budget decision. Measured savings from scoping one payload, and why a thin result is more dangerous than a fat one.
A verification step that can pass while the bug is present is not a verification step
Static checks read what was written; the defect lives in what was drawn. Four artifacts that passed every gate this session and failed the moment someone looked.
Write down what you did to an external service
A create-only API cannot answer whether a thing exists. The local ledger becomes the sole source of truth, and one write's ordering decides if it works.
Three kinds of code I stopped letting an agent write
An honest negative result: three categories where generated code cost more to verify than to author, and the property they share that predicts it.
Teaching an agent when to stop
Enumerated halt classes make stopping checkable; a mood does not. The forbidden reasons that stall a drive, and the undeclared halt I made anyway.