arena

Files (1)

• SKILL.md

SKILL.md (first lines)

# Arena

Open a todolist 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 `.claude/rules/mstack-models.md`. If the file or that line is missing, default to one each on `fable`, `opus`, and `sonnet`, three seats, one per tier. An `auto` or `inherit-parent` entry in this line or the cross-judge line means the parent model, so omit `model` for it. If the Agent tool rejects a configured entry, run that seat on the default for its position and say so. Set `effort` per **Effort by tier** in [`poteto-mode`](../poteto-mode/SKILL.md). Never draw the whole field from a single tier; tier spread is what keeps the candidates from converging on the same shape. 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.
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.

## Phase B: Fan out

Spawn all N subagents in one message, each with the task, the 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.

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 the cross-judge model from the `arena cross-judge` line in `.claude/rules/mstack-models.md`. If the file or that line is missing, use `fable`. Prefer a different **tier** from the parent's, and always an independent context. A judge that shares the parent's context inherits the parent's blind spots, which is the failure a cross-judge exists to prevent. Spawn one judge subagent on that model, instructed to read and score only; the parent applies any change. It sees the rubric and the candidates by 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, or the judge sees partial or empty outputs and reports them as dropouts.

## 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.

From MatrixMagician/mstack, MIT.

More from MatrixMagician/mstack