Skip to content
Back to skills

Setup

ASecurity

Bootstrap the vteam proof-of-done harness in the current repo — grade it first (npx vteam-harness audit), install (npx vteam-harness init), verify (npx vteam-harness doctor). Use when the user asks to set up vteam, add proof-of-done gates, or install the virtual AI team.

  • 6 stars
  • 0 votes
  • 0 copies
  • 4 views
  • Added September 2, 2026
ai-agentsrustgonodegit

Works with

  • claude code
  • cursor
  • cli

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add connorpham/vteam-harness --skill setup --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Setup?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Setup
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/connorpham-setup/badge)](https://www.skillsdirectory.com/skills/connorpham-setup)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
description: Bootstrap the vteam proof-of-done harness in the current repo — grade it first (npx vteam-harness audit), install (npx vteam-harness init), verify (npx vteam-harness doctor). Use when the user asks to set up vteam, add proof-of-done gates, or install the virtual AI team.
---

# vteam setup — proof-of-done for AI agents

First explain, in substance, these 5 lines and nothing grander:

1. Agents say "done"; vteam makes them prove it — machine gates that actually go red.
2. Every checking gate ships a `--selftest` mutation proof (green fixture + red mutations): the gate itself is tested, not trusted.
3. Evidence is pinned to commits — claims carry command + real output written to files, not chat scrollback.
4. `.vteam/manifest.json` guards updates (user edits are never silently clobbered) and the secret scan fails closed.
5. The same gates drive Claude Code, Cursor, Windsurf, Codex, and Copilot — one team, any agent tool.

Then run the funnel in order. Show the user each command's REAL output; never claim a step succeeded without it.

## 1. Grade the repo first

```
npx vteam-harness audit
```

Scores the repo 0-100 on proof-of-done before changing anything, so the user sees the gap the install closes. If the installed vteam-harness predates the audit command (the CLI prints usage without `audit`), say exactly that and continue with step 2 — do not fake a score.

## 2. Install

```
npx vteam-harness init
```

Interactive; `--yes` accepts defaults, and flags like `--key PROJ --profile node --tools claude-code` skip prompts. Ask before running it: it writes `vteam.config.yaml`, `.vteam/` (gates + manifest), `.claude/` (skills + SessionStart hook), `.githooks/pre-push`, CI workflow, and docs skeletons. It refuses to run twice — an existing install updates with `npx vteam-harness update`.

## 3. Verify — nothing is trusted untested

```
npx vteam-harness doctor
```

Preflight + install integrity against the manifest + every installed gate's `--selftest`. Green doctor is the definition of installed; anything red gets reported to the user verbatim, not smoothed over.

When doctor is green, tell the user to start a workday with `/team`.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.