cameronjlarsen/engineering-toolkit/swarm

Swarm

Fan out N parallel workers. They may cover separate slices, race the same brief, or mix both. The parent waits, aggregates, and returns one report.

Dispatch contract. Resolve each worker Route through `provider-dispatch.md`. The parent starts the lanes; workers never route themselves. On Codex, resolve remaining Claude tool names via `codex-tools.md`.

Start

Open a todolist with one entry per phase before launching anything.

1. Frame
2. Fan out
3. Aggregate
4. Report

Phase A: Frame

1. State the done predicate and the artifact or report the swarm must return.
2. Choose the shape. Partition into slices, race N workers on identical briefs, or mix both. For a race or mixed shape, declare first pass, rank all, or best-of before spawning.
3. Set N from the user or derive it from the shape. N is total workers, not the number that run at once.
4. Pick the worker descriptor from the swarm workers line in the active model sheet (see `t3-tools.md`). If the sheet or that line is missing, use inherit-parent. For auto or inherit-parent, omit the Route so the workers run on the parent model. For a model race, the operator names each arm's descriptor up front; the single swarm workers entry is never split into arms. With no configured arms, the race is a same-model repeat; say so.
5. Give each worker its own writable output when it writes. When workers verify or measure commits, each brief names the exact SHAs. A measurement brief also names the method (sample count, what one sample is, order). The worker records both in its result.

Phase B: Fan out

Start all N workers in one fan-out phase through provider dispatch. Workers are background subagents with retained task handles. Never use Claude's foreground Bash path for a long worker. Every writer runs in its assigned worktree or output directory. Isolation comes from those paths, not the provider.

When a worker must start from a non-default branch, check that branch out in the worker's own worktree and name the worktree path in its brief.

Every brief stands alone. Include the goal, scope, exact slice or race arm, how to verify, and what to report. Reports use PASS, ISSUES, or BLOCKED with evidence. A worker that can prove a defect reports ISSUES and lists every issue it can prove, not only the first.

If a worker drops out, proceed with N-1 and note the provider, model, and failure. Never substitute another provider silently.

Phase C: Aggregate

Read the terminal results. Drop a result that does not record the SHAs and method its brief names, and rerun that worker once. After a second miss, record a gap. A gap does not count as a pass. For coverage, every required slice needs a result. For a race, apply the selection rule declared up front. Use first pass, rank all, or best-of. Do not paste raw worker dumps.

Keep a compact result table, one-line evidenced issues, and explicit gaps or dropouts.

Phase D: Report

Return one consolidated in-chat report with a result table, issue one-liners, gaps or dropouts, and the race rule when used.