Topic
Tool-call design inside agent harnesses, including blocking versus asynchronous tools and how results re-enter model context.
A post arguing the Model Context Protocol was built for 2024-era models pushed a 165-point Hacker News thread into a fight about token cost. The post reports no benchmark, so the measurement is left to your own agent.
Reality
The MIT-licensed framework gives agents one command grammar for Salesforce, ServiceNow, DocuSign and Agentforce. The safety metadata a host checks before running a command comes from provider authors the project leaves unnamed.
Reality
The wire format an internal tool returns to a coding agent is a line item on the inference bill. But the headline saving in one Sonar CLI comparison depends on which JSON baseline you measure against, and on your tokenizer.
Reality
The proposal swaps DOM inference for tools the page registers itself, which is sound engineering and also a standing maintenance bill for whoever owns the state machine. No shipping browser is named in the write-up.
Reality
The author of Doco argues that browse, search, cite and safe-edit operations over a maintained corpus do most of the work teams expect from embeddings, and that the pipeline adds a copy which drifts from the source.
Reality
OpenAI merged an async developer-message tool into the public Codex repository, so the agent no longer blocks on your answer. Nothing in the change stops it from coding past your decision.
Reality