Skip to content
Back to skills

Agent Harnesses

ASecurity

Operate and configure Claude Code, Codex, Copilot, Grok, and OpenCode.

  • 10 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 23, 2026
ai-agentsgit

Works with

  • claude code
  • cli
  • mcp

Security analysis

A100/100

Pro scans all 11 files and shows the line behind each finding

Scanned October 6, 2026

npx -y skills add fmind/dot --skill agent-harnesses --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Agent Harnesses?

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

Security grade badge for Agent Harnesses
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/fmind-agent-harnesses/badge)](https://www.skillsdirectory.com/skills/fmind-agent-harnesses)

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
---
name: agent-harnesses
description: "Operate and configure Claude Code, Codex, Copilot, Grok, and OpenCode."
license: MIT
metadata:
  kind: collection
  author: Médéric HURIER (Fmind)
  source: github.com/fmind/dot/tree/main/skills/agent-harnesses
  created: "2026-09-16"
  updated: "2026-10-06"
---

# Agent Harnesses

Operate the selected coding-agent host without conflating its configuration, permissions, or lifecycle with another host. Antigravity has its own [agy](../agy/SKILL.md) owner.

## Workflow

Every host follows these steps; read only the matching guide for its specifics and required resources.

1. **Inspect the installed contract when it matters**: run the host's `--version` and `--help` for unfamiliar or version-sensitive features; identify the surface (CLI, app, IDE, or cloud), workspace, configuration scope, and the session to start or resume.
1. **Refresh evolving details**: read the relevant official page and changelog before relying on flags, settings, models, or feature availability. Compare with installed help, which can omit supported flags; report version gaps before applying a newer recipe, and keep release-specific details upstream.
1. **Use the documented interface** for the session, configuration, or extension. Resolve configuration precedence and permission scope before changing execution behavior; hooks and extensions execute code.
1. **Verify** the result, resulting artifacts, and the host's own instruction or skill discovery. Use [agent-project](../agent-project/SKILL.md) for shared instruction layout and [mcp-setup](../mcp-setup/SKILL.md) for MCP registration.

## Task guides

<!-- guides:start -->

- [claude](references/claude/GUIDE.md): Claude Code sessions and customization.
- [codex](references/codex/GUIDE.md): Codex sessions and customization.
- [copilot](references/copilot/GUIDE.md): Copilot sessions and extensions.
- [grok](references/grok/GUIDE.md): Grok Build sessions and configuration.
- [notifications](references/notifications.md): Diagnose or change managed turn notifications and their hooks per host.
- [opencode](references/opencode/GUIDE.md): OpenCode sessions and provider configuration.

<!-- guides:end -->

Files in this skill

  • SKILL.md2.9 KB
  • references/claude/GUIDE.md2 KB
  • references/claude/references/features.md2.5 KB
  • references/codex/GUIDE.md2 KB
  • references/codex/references/features.md2.6 KB
  • references/copilot/GUIDE.md2 KB
  • references/copilot/references/features.md3.4 KB
  • references/grok/GUIDE.md1.8 KB
  • references/grok/references/features.md2.9 KB
  • references/opencode/GUIDE.md4.4 KB
  • references/opencode/references/features.md2.3 KB

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…