Confer — cross-model consultation with resumable threads

You talk to a peer model through threads: open one with a question, the peer's session id is stored, and any later round — today or next week, from any host session — resumes the same peer-side context. All mechanics live in scripts/confer.mjs (single source of truth); never assemble raw claude -p / codex exec calls yourself.

scripts/confer.mjs open <provider> [-t name] <prompt|->   # start thread (claude|codex)
scripts/confer.mjs reply <thread> <prompt|->              # continue with full peer-side context
scripts/confer.mjs all <prompt|->                         # fan one question out to every provider (concurrent)
scripts/confer.mjs list | show <thread>                   # registry / transcript
scripts/confer.mjs doctor [--live]                        # health check (run --live once after install)

Pass - as the prompt and pipe stdin for anything long or containing quotes.

A round can take minutes. When you expect a long consultation and have other work, run the call in the background and pick the reply up when notified — never relay a peer through a subagent: the peer's own words must reach the user undiluted. Each ← transcript header records which model answered (and cost/tokens where the CLI reports them).

Steps

1. Resolve the target. Which provider, and new thread or continuation? Prefer a peer that is not your own model family — a same-model consult is not an independent second opinion. When the user says 继续/上次/"what does it say now", run list and match — never open a fresh thread for what is semantically round N of an old one. Name threads you expect to revisit (-t zhang-pe-review); let one-shots auto-name. Done when: provider + thread decided.