gauntlet SKILL.md skillA

gauntlet from tomw1808/flash-agents is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add tomw1808/flash-agents/gauntlet@git:20260916.5099f50

Pins git:20260916.5099f50 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
tomw1808/flash-agents
Path
skills/gauntlet/SKILL.md
Versions recorded
3
Latest version
git:20260916.5099f50, committed Sep 16, 2026
First version recorded
Sep 16, 2026
Size
23,303 bytes
Repository stars
7
License
MIT

Orchestrate cheap parallel builder and critic subagents toward a runnable quality bar. The orchestrating (frontier) model frames the task, filters findings, and judges finalists in a blind A/B duel; smaller models do the building, critiquing, and revising. Use for tasks whose result can be checked by running something. Do not use for design decisions or vague goals.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Gauntlet: frontier-model orchestration, cheap-model execution

> **Status: shipped with the `flash-agents` plugin, and still young.** Installing the plugin
> makes this available as `/flash-agents:gauntlet`. It assumes no helper script: every
> mechanical step is written as plain shell/git, so any agent harness can follow it — and it has
> been reasoned through more thoroughly than it has been run end to end. Read §10 (open
> questions) before handing it a long or expensive job, and run it on a task whose bar you can
> actually execute.

## The idea in one paragraph

You, the orchestrator, do the brain work: write the spec, write the bar (runnable acceptance
checks), split the work, filter what critics find, and pick winners. Cheap subagents do the
legwork: several *different* attempts in parallel, adversarial critique, and revisions. The
filesystem, not your context, carries state between rounds. Each round a challenger must beat
the current champion in a **blind side-by-side comparison**. The loop ends when the champion
keeps winning, when the budget runs out, or when progress stops.

Adapted from the gauntlet-loop pattern (named and fetchable bar, fresh-context critic, blind
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add tomw1808/flash-agents/gauntlet@sha256:4de81c09b62f0d6a

The line at the top of this page pins the label git:20260916.5099f50, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_m44n63ihtufhajk7.svg)](https://markdownregistry.com/a/art_m44n63ihtufhajk7)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 3 versions of this file, all committed on Sep 16, 2026.

versioncommittedcommitsizeaudit
git:20260916.5099f50 latest2026-09-16 5099f50 23,303 BA view · diff
git:20260916.bd368852026-09-16 bd36885 23,141 BA view · diff
git:20260916.d556fd42026-09-16 d556fd4 21,915 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (23303 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

tomw1808/flash-agents · 7 stars · license MIT · pushed 2026-10-31 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_m44n63ihtufhajk7
GET https://markdownregistry.com/api/v1/resolve?ref=tomw1808/flash-agents/gauntlet
GET https://markdownregistry.com/api/v1/blob/4de81c09b62f0d6a666b853645f1eeec88b0a52a83335faf2ec9bbf706fe1431

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from tomw1808/flash-agents

delegate SKILL.md skill
tomw1808/flash-agents · skills/delegate/SKILL.md · Run a goal through flash workers instead of Claude subagents: frame it, cut it into coherent slices, dispatch with…
git:20261028.d7ec236 · audit A · 7 stars
setup SKILL.md skill
tomw1808/flash-agents · skills/setup/SKILL.md · Check and finish the Flash Agents prerequisites — the harness, Ollama, the default and allowed models, and the service…
git:20261028.d7ec236 · audit A · 7 stars

Every file in tomw1808/flash-agents

Other files named gauntlet

gauntlet SKILL.md skill
aiskillstore/marketplace · pending/aakashh242/gauntlet/SKILL.md · Evidence-driven adversarial review and root-cause repair of code, designs, APIs, infrastructure, tests, and agent…
v1.0.0 · audit A · 452 stars
gauntlet SKILL.md skill
oprogramadorreal/optimus-claude · skills/gauntlet/SKILL.md · Runs a Gauntlet Loop: turns an ambitious goal into a builder/critic prompt judged against a concrete quality bar…
git:20260925.7013524 · audit B · 75 stars
gauntlet SKILL.md skill
bountyoperator/bounty-operator · skills/gauntlet/SKILL.md · Runs a security finding through Bounty Operator's eight pre-submission stages in order (scope, provenance, prior art…
v0.9.21 · audit A · 5 stars
gauntlet SKILL.md skill
bountyoperator/bounty-operator · web/public/.well-known/agent-skills/gauntlet/SKILL.md · Runs a security finding through Bounty Operator's eight pre-submission stages in order (scope, provenance, prior art…
v0.9.21 · audit A · 5 stars

Browse by kind, by grade A, or by owner.