| name | Gauntlet |
| trigger | /gauntlet |
| description | Gauntlet loop — multi-agent quality harness: a lead sets a real reference bar (a named product, test suite, or baseline to match or beat), splits the goal into gradeable units, builders produce artifacts in clean context, and blind fresh-per-round critics grade each artifact against the bar with evidence; failures loop until every unit passes, two rounds stall, or budget runs out. NOT /challenge (verifies factual claims); NOT /review (one-pass verdict) — this rebuilds until the work beats something real. |
| agent | gauntlet-lead |
| arguments | [{"name":"goal","description":"What to build or improve until it beats the bar (e.g., \"landing page that beats linear.app's\", \"CSV parser matching papaparse's test suite\")","required":false},{"name":"--bar","description":"The exemplar to match or beat — a reference product/screenshot, test suite + threshold, model doc, or measured baseline","required":false},{"name":"--budget","description":"Max rounds (default 5)","required":false}] |
Triggers the gauntlet-lead subagent in a forked context.
Multi-agent quality harness (Matt Shumer's gauntlet-loop technique): the LEAD sets a real
reference bar and splits the goal into independently gradeable units; BUILDERS produce artifacts
in clean context; BLIND CRITICS — a fresh context per unit per round, seeing only the artifact,
the bar, and the exemplar — grade with evidence; failures loop back until every unit passes, two
rounds stall, or the budget runs out.
The two inviolables: the agent that builds never grades its own work, and a critic that saw a
previous draft never grades the retry. Prior critiques go to the builder as fix input, never to
the next critic.
The bar must be real: a named product screen, a test suite + threshold, a model doc, a
measured baseline — written to docs/gauntlet/BAR_<slug>.md before round 1. "Make it amazing"
gets BLOCKED: no real bar. Aspirational bars are legitimate; uncomparable ones are not.
Protocol: agents/shared/GAUNTLET_LOOP.md — including boundaries vs /challenge (fact-checks
claims), Fix-Verify (closes known defects), and /design-iterate (token conformance against our
own spec — cheaper when the bar is internal).
Outputs:
docs/gauntlet/BAR_<slug>.md — exemplar, per-criterion checks, budget
docs/gauntlet/GAUNTLET_<slug>.md — round log, PASS evidence, which exit rule fired, below-bar residuals