coder/xum is a desktop application for running coding agents in isolated, parallel workspaces, either locally, through Git worktrees, or on remote servers over SSH. It helps developers manage multiple agents, models, code changes, reviews, costs, and agent status from one interface. The catalogue skills, instructions, agents, settings, and command define workflows for using Xum.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add coder/xum --skill dev-desktop-sandboxgit clone --depth 1 https://github.com/coder/xumWrote 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/coder/xum/dev-desktop-sandbox)<a href="https://agentmods.dev/skills/coder/xum/dev-desktop-sandbox"><img src="https://agentmods.dev/badge/skills/coder/xum/dev-desktop-sandbox/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/coder/xum/dev-desktop-sandbox"><img src="https://agentmods.dev/badge/skills/coder/xum/dev-desktop-sandbox.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
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 →
- high Privilege Escalation · line 124 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
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 | $0.00024 | $0.01292 |
| Opus 5.5 | $0.00010 | $0.00517 |
| Sonnet 5.5 | $0.00005 | $0.00258 |
| Haiku 4.5 | $0.00002 | $0.00129 |
Grade A, and why
dev-desktop-sandbox 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 1mo 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.
The source is not reproduced here
Licensed AGPL-3.0
The repository is licensed AGPL-3.0, which this catalogue does not treat as permission to reproduce the file. Read it at the source.
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.
- 1mo ago First seen · 125 lines · 24 tokens per session scan A 26beb1d89946
dev-desktop-sandbox is a skill published in the GitHub repository coder/xum (2,054 stars, last pushed today), licensed AGPL-3.0. It adds 24 tokens to every session and 1,292 once invoked, about $0.0001 per session on Opus 5.5. 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
novel-standard
Standard writing workflow in which the main Agent drafts and revises while the built-in general-purpose SubAgent reviews, balancing quality and speed.
web-research
Use when a task needs current public-web facts, comparisons, fact-checking, sources, or a multi-source synthesis and both websearch and webfetch are available.
configuration
Use when a Project Agent reads or changes lore indexes and group memberships, managed presets, game modules, automations, Skills, or Agent profiles through configread and configapply.
extension-development
Develop, repair, and validate a Denova plugin or playable game in its source Project.
novel-lite
Fast continuation, exploratory drafts, and low-latency prose generation performed directly by the main Agent without review or revision subprocesses.
chapter-illustration
Generate one spoiler-free illustration for the current or specified chapter and save it under workspace assets/writing.