arena

Arena

A skill named in bold or in backticks, or run as /<name>, is the file ../<name>/SKILL.md beside this skill's directory, and a principle named without its prefix is ../principle-<name>/SKILL.md. Read the file and follow it. Most pstack skills are user-only, so your skill tool cannot load them. Every delegate_task call follows the defaults in the Subagents section of ../poteto-mode/SKILL.md.

Open a task list (your harness's task tool, or a checklist file at ~/.local/state/pstack/checklists/<repo>-<branch>.md when it has none) with one entry per phase (Frame, Fan out, Cross-judge, Pick, Graft, Verify) before launching anything.

Phase A: Frame

Every candidate gets the same prompt, so the prompt is the contract.

1. State the artifact each candidate is producing.
2. Derive the rubric, 3-6 concrete gradeable criteria for what success looks like on this task. It is the picker's tool in Phase D. Candidates only see the task.
3. Pick the runners. Use the arena runners line in ~/.config/pstack/pstack-models.md, one seat per entry. If the file or that line is missing, default to one seat on the parent thread's model and one on the strongest model of each other provider that orchestrator_capabilities reports with canRunChildTask, up to three seats. Say so and suggest /setup-pstack. An entry <providerInstanceId>/<model> (<effort>) sets the seat's provider, model and the effort option id orchestrator_capabilities lists for that model. An auto or inherit-parent entry in this line or the cross-judge line means the parent model: omit the whole target for a delegate_task seat, and omit modelSelection for a t3_thread_launch candidate so it inherits. If T3 rejects a configured entry, run that seat on the closest model of the same provider that orchestrator_capabilities lists and say so. Spawn more runners when the arena covers multiple design directions. Use the same model N times when the work is generation-bound rather than judgment-sensitive. Pick each subagent's reasoning level and fast or standard speed for the job, per the Subagents section of the poteto-mode skill.
4. Give each candidate its own output location (a git worktree where possible, otherwise /tmp/arena-<slug>/candidate-<n>/), per the separate-before-serializing-shared-state principle skill. A candidate that writes code gets its own worktree. Commit the base first, because uncommitted edits do not reach a new worktree.

Phase B: Fan out

Spawn all N candidates in one turn, each with the task, the absolute path to the shared grounding, its own output path, and instructions to produce the artifact and a short rationale naming the alternatives it considered and what it rejected. A candidate that writes code is a t3_thread_launch thread titled arena <slug>: candidate <n>, with workspaceStrategy: {"type":"worktree","baseRef":"<base commit>","branch":"arena-<slug>-<n>","startFromOrigin":false}, modelSelection: {"instanceId": ..., "model": ..., "options": ...} from its seat, and the brief in message. Wait on it with t3_thread_wait and take its worktree path from t3_thread_read. Any other candidate is a delegate_task child with mode: "async", the same title, target from its seat, and its /tmp output path. Its completion notification wakes you.

If a candidate produces no output, proceed with N-1 and note the dropout.

Phase C: Cross-judge

After all Phase B candidates complete, choose one model from the arena cross-judge pool line in ~/.config/pstack/pstack-models.md. If the file or that line is missing, choose from the strongest model of each provider that orchestrator_capabilities reports with canRunChildTask. Prefer a different model family from the parent's. Spawn one judge on that model as a delegate_task child with mode: "async", titled arena <slug>: cross-judge, its brief saying read-only: inspect only, no writes, no git commands. It sees the rubric and the candidates by absolute path label, scores each criterion, and recommends a base with rationale. It runs in parallel with your reading in Phase D. Don't spawn it while candidates are still writing.

Phase D: Pick a base

Read every candidate end to end before picking. Score each against the rubric criterion by criterion, not on holistic feel, and compare with the cross-judge. Agreement on the base confirms the pick. Disagreement means one of you is biased or the rubric was ambiguous. Read both rationales before deciding.

Pick the base a future maintainer can extend most easily without breaking invariants. When two feel tied, prefer the cleaner boundary or smaller API, per the Laziness Protocol.

Phase E: Graft

Walk each losing candidate once more for what is worth porting into the base, usually one or two things per candidate, not most of it. Fold each graft in by hand, per the redesign-from-first-principles principle skill, so the result stays coherent under one mental model. Don't paste mechanically.

If the candidates converge on one shape, that is strong agreement. Note it and ship the consensus shape. No graft is needed. If they wildly diverge, Phase A was under-specified. Reframe and re-run rather than averaging the divergence.

Phase F: Verify

Verify the synthesized artifact like any other output, per the prove-it-works principle skill. If verification finds a problem the arena missed, either Phase A was wrong (re-frame and re-run) or a candidate caught it and you missed the graft (go back to Phase E). Don't paper over it.

Outputs

One synthesized artifact, with a short synthesis note alongside naming the base and why, the cross-judge's verdict, each graft with its source candidate, what was rejected and why, any convergence or dropouts, and the verification result.