codex-host: Skill for Claude Code

.agents/skills/codexhost-add-harness/SKILL.md

codexhost-add-harness is a skill for Claude Code, Codex from BytePioneer-AI/codex-host. It costs 66 tokens per session (2,239 once invoked), scanned A, original, MIT.

A guide for adding a new external coding agent, called a harness, to the codexhost system. It explains how that agent should connect to the system's existing shared interfaces and capabilities.

In plain words
What is it for?
Use it to plan, implement, review, or complete a harness adapter and to check support for identity, snapshots, resume, fork, rollback, permissions, and related lifecycle features.
Why use it?
It helps avoid building a one-off integration that does not fit the host system. It also makes unsupported features explicit instead of silently failing.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions subagents; installed under .agents/ (shared by several agents).

This is BytePioneer-AI/codex-host's own configuration. It tells Claude Code and Codex how to work on codex-host itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything codex-host configures →

About the project

CodexHost is a desktop application that lets users run coding-agent harnesses such as Pi and Claude Code inside Codex Desktop. It is for people who want to choose among multiple agents and have them delegate tasks to one another while staying in one workspace.

BytePioneer-AI/codex-host · 2,209 stars · on GitHub

Reuse

Borrowing it

Nothing to install: this file belongs to BytePioneer-AI/codex-host. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/BytePioneer-AI/codex-host/main/.agents/skills/codexhost-add-harness/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/BytePioneer-AI/codex-host

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for codexhost-add-harness

README.md
[![agentmods](https://agentmods.dev/badge/skills/bytepioneer-ai/codex-host/codexhost-add-harness/github.svg)](https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-add-harness)
Your own site
<a href="https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-add-harness"><img src="https://agentmods.dev/badge/skills/bytepioneer-ai/codex-host/codexhost-add-harness/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for codexhost-add-harness

Your own site · 80×15
<a href="https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-add-harness"><img src="https://agentmods.dev/badge/skills/bytepioneer-ai/codex-host/codexhost-add-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 66 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,239 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00066 $0.02239
Opus 5 $0.00033 $0.01120
Sonnet 5 $0.00013 $0.00448
Haiku 4.5 $0.00007 $0.00224

Measured 5d ago against content hash a2a40197ecb6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

codexhost-add-harness scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/codexhost-add-harness/SKILL.md · 119 lines

How it starts

The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.

新增 codexhost Harness 插件

目标:交付一个通过公共契约工作的独立插件,而不是在 Host 中增加一套 Harness 专用流程。插件是交付与加载单元,HarnessAdapter / HarnessSession 是运行时接口;不要求继承基类,也不要求使用 ACP。

本文代码路径以仓库根目录为基准,references/ 链接相对于本 Skill。接口签名、字段限制和当前行为以源码为准;参考文档是实现与验收清单,不是第二份接口定义。

当前架构边界

Host
├─ 发现 Manifest、读取显式启用配置、调用插件工厂
├─ 通过 HarnessAdapter / HarnessSession 执行业务
└─ 拥有 Thread 映射、持久化事务、委派和 Desktop 协议投影
             │ 公共契约
             ▼
插件
├─ Manifest:身份、版本、展示信息、入口
├─ 工厂:把 Host Context 转成原生构造参数
├─ HarnessAdapter:检查、打开 Session、关闭共享资源
└─ HarnessSession:执行、输出、状态、历史、关闭
             │ 插件内部
             ▼
原生 SDK / RPC / 服务 / CLI
  • 七个既有 Harness 和用户插件共用动态加载器,Host 不静态依赖具体 Adapter 包。新 ID 使用共享插件路由。
  • 公共工具仍可复用:例如 harness-discovery 实现搜索机制,插件声明命令名、环境变量和安装目录策略。
  • 完整产品去专属化尚未完成:Renderer 仍有固定 Agent 名单、配置字段和展示接线;旧路由、部分恢复策略、Credits、导入、Broker 与远程配置仍有历史特例。插件加载成功不等于自动出现在 Desktop。
  • 既有特例是兼容负担,不是新 Harness 的实现模板。公共契约无法表达真实需求时,记录缺口并在当前任务范围内设计公共扩展;不得通过新 Harness 专用 Host 分支绕过契约。

1. 确定交付范围和能力

从用户请求判断范围;只有会影响实现边界的歧义才询问,不默认扩展为全产品改造。

交付范围 必须完成 不自动包含
插件后端 原生适配、Manifest/工厂、可运行依赖、隔离目录加载、公共行为测试 仓库预装、Desktop UI
预装发行 插件后端 + Workspace 构建、预装清单、独立 Bundle、许可与发行验收 Renderer 自动发现
Desktop 产品接入 插件后端 + Picker、配置、恢复、展示及 UI 验收 必须成为预装插件

普通持久化 Thread 和完整委派要求可写 resume。原生系统不支持时,可以交付明确受限的后端,但不能把它称为完整产品接入;当前没有自动 ephemeral Thread 降级。

先读取以下权威入口:

  • packages/harness-adapter/src/text-session.ts:Adapter、Session、命令、事件和可选接口。
  • packages/harness-adapter/src/plugin.ts:工厂和 Context。
  • packages/shared-contracts/src/harness-models.ts:Catalog、能力和检查结果。
  • packages/shared-contracts/src/harness-plugins.ts:Manifest、描述与启用配置。
  • docs/harness-plugin-runtime.md:已实现边界、安装、信任和发行规则。

核对目标 Harness 的当前原生接口、版本、认证和运行方式。实际可行时优先原生 SDK、RPC 或服务接口;只有原生接口不可用或确有理由时使用 ACP,并记录能力差异。不要以 CLI 名称相似推断协议兼容。

使用 实现导航 按传输和能力选参考,不整包复制。输出简短计划:

交付:插件后端 / 预装发行 / Desktop(可组合)
原生接口与版本:……
支持:create、resume、Turn、取消、历史、工具、Question……
不支持或受限:Fork、权限仅创建时设置……
待验证:……(不能当作不支持,也不能声明已实现)
参考模块:……
预计修改:插件目录;其他位置逐项说明原因

Read the full file on GitHub · 119 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago Changed a2a40197ecb6
  2. 6d ago Changed · +54 lines · -17 tokens per session d8bfc580ace1
  3. 12d ago First seen · 65 lines · 83 tokens per session scan A 3f8ee2945c6f

Subscribe to this mod's changes

codexhost-add-harness is a skill published in the GitHub repository BytePioneer-AI/codex-host (2,209 stars, last pushed today), licensed MIT. It adds 66 tokens to every session and 2,239 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.