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.
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.
curl -O https://raw.githubusercontent.com/BytePioneer-AI/codex-host/main/.agents/skills/codexhost-update-impact-audit/SKILL.mdgit clone --depth 1 https://github.com/BytePioneer-AI/codex-hostWrote 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.
[](https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-update-impact-audit)<a href="https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-update-impact-audit"><img src="https://agentmods.dev/badge/skills/bytepioneer-ai/codex-host/codexhost-update-impact-audit/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.
<a href="https://agentmods.dev/skills/bytepioneer-ai/codex-host/codexhost-update-impact-audit"><img src="https://agentmods.dev/badge/skills/bytepioneer-ai/codex-host/codexhost-update-impact-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 134 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium Excessive Agency · line 185 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00060 | $0.02081 |
| Opus 5 | $0.00030 | $0.01040 |
| Sonnet 5 | $0.00012 | $0.00416 |
| Haiku 4.5 | $0.00006 | $0.00208 |
Grade A, and why
codexhost-update-impact-audit 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.
codexhost update impact audit
Audit before changing codexhost. Produce a verdict backed by bundle and live-Renderer evidence. Apply a fix only when the user explicitly requests one.
可先运行 npm run audit:codex-desktop 辅助检测；该命令不能替代下面的语义对比、ownership 追踪、真实 Renderer 探测和必要的行为验证，后续步骤仍须继续执行。
Guardrails
- Read the repository
AGENTS.mdand recordgit status; preserve unrelated dirty-worktree changes. - Treat Codex Desktop private DOM, React state, and main-process services as versioned contracts.
- Use semantic attributes, API shape, ownership, and observed relationships as contracts. Bundle hashes, asset names, minified identifiers, localized labels, private CSS classes, credentials, prompts, and full payloads are evidence only.
- Start with a read-only inspection of an existing codexhost-controlled Desktop. Use a controlled launch or mutate Renderer state only when the required boundary cannot otherwise be verified.
- Store only sanitized evidence under ignored
.codexhost/update-impact/. Never persist Thread IDs, request IDs, prompts, transcripts, tokens, credentials, URL query/hash values, or full DOM snapshots. - Do not change production code during an audit. Do not claim a live or routing result unless that exact check ran.
1. Establish the comparison chain
Record the installed executable, Desktop version/build, Chromium version, and app.asar hash. Determine these distinct versions when possible:
- reviewed baseline: the last version whose audit passed;
- direct predecessor: the version immediately before the installed build;
- current: the installed build under review.
Use the reviewed baseline for the compatibility decision. Use a direct predecessor only to localize the newest change. Never silently substitute a much older bundle for either role.
Search, in order, for prior evidence under .codexhost/update-impact/, Sparkle installation caches, compatibility fixtures, and other complete local installations. The official Sparkle appcast may establish release order and download locations, but downloading a full application is optional and must not block the audit when live evidence and a reviewed baseline are available.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 11d ago First seen · 189 lines · 60 tokens per session scan A a45c3813032e
codexhost-update-impact-audit is a skill published in the GitHub repository BytePioneer-AI/codex-host (2,141 stars, last pushed today), licensed MIT. It adds 60 tokens to every session and 2,081 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.
Other skills, from other repositories
smell
Detect software architecture bad smells, algorithmic complexity hotspots, and anti-patterns in a codebase. Produces a detailed markdown report identifying violations of architectural principles, design patterns, code quality, and performance complexity. Triggers on: smell, code smell, architecture smell, find…
surf
Control Chrome browser via CLI for testing, automation, and debugging. Use when the user needs browser automation, screenshots, form filling, page inspection, network/CPU emulation, DevTools streaming, or AI queries via ChatGPT/Gemini/Perplexity/Grok/AI Studio.
performance-optimizer
Agents should invoke this skill for slow code, high CPU/memory, latency, large data processing, algorithmic complexity, profiling plans, benchmarks, or optimization requests. Profiles first and weighs trade-offs before changing code.
network-diagnostics
Agents should invoke this skill for connectivity, DNS, Pi-hole, port reachability, routing, firewall reachability, TLS/network timeouts, or service access failures. Provides structured network troubleshooting commands and interpretation.
refactoring-advisor
Agents should invoke this skill for refactors, code smells, migrations, duplication removal, module splitting, API cleanup, or restructuring plans. Emphasizes small safe steps, behavior preservation, and verification after each change.
bug-reporter
Agents should invoke this skill when defects, regressions, failed tests, unexpected behavior, or spec mismatches are found. Produces structured reproducible bug reports with severity, evidence, environment, and actionable next steps.