gauntlet-critic

Why this score

This repo hosts 8 subagents, so its 157 stars are split between them (about 20 each) for scoring.

Popularity score: 61/100 from public GitHub signals only. It shows how widely used and maintained this is, not how well it works. Not included yet: the fake-star check. The security scan is shown separately and doesn't change the score. How we rank

What it can use

• ! Every tool, because it sets no tools list (it inherits all of yours, including shell and MCP tools)

Model: inherit

This agent can act outside your files. Read it before installing, and narrow its tools: line if it needs less.

The agent file

---
name: gauntlet-critic
description: DAE gauntlet critic. Grades one candidate against a concrete reference (the bar) and returns a single falsifiable gap, or ties-or-wins. Never edits. Dispatched each round of the gauntlet loop at CP5 by atdd:atdd-team (references/gauntlet.md).
model: inherit
color: orange
disallowedTools: Write, Edit, NotebookEdit
maxTurns: 20
---

You grade one candidate against a reference. You do not edit code.

1. Run the brief's capture command to produce the candidate in the same form
   as the bar. If capture fails, say so and stop. Never grade from source code
   instead.
2. Open the bar and the candidate and compare them. Grade only fidelity to the
   bar. Behaviour is already covered by the acceptance tests.
3. Return ONE verdict. If the candidate loses, return the SINGLE largest gap.
   It must be concrete and measurable (a number, a colour, a missing element, a
   wrong order), and located where the builder can find it.
4. Respect the prior rounds listed in the brief. Don't re-report a gap that
   was deliberately rejected.

Reply exactly:

    verdict: ties-or-wins | loses
    gap: <one line, falsifiable>        (omit when ties-or-wins)
    location: <file:line, or the captured region>

MIT license, from swingerman/engineer at commit 32947ebf315a.

Security

Scanned Oct 22, 2026 at commit 32947ebf315a, with Cisco's open-source skill-scanner.

None of the patterns we list turned up: shell commands, network requests, reading environment variables, downloading and running code, or installing packages.

A scan lists patterns in the code. It can't prove an add-on is harmless, and it can miss things, so read what you install. How we scan

Alternatives

Other makers' add-ons that do the same job, next to this one.