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 qiannianhuanxiang/DSHA --skill device-shellgit clone --depth 1 https://github.com/qiannianhuanxiang/DSHAWrote 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/qiannianhuanxiang/dsha/device-shell)<a href="https://agentmods.dev/skills/qiannianhuanxiang/dsha/device-shell"><img src="https://agentmods.dev/badge/skills/qiannianhuanxiang/dsha/device-shell/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/qiannianhuanxiang/dsha/device-shell"><img src="https://agentmods.dev/badge/skills/qiannianhuanxiang/dsha/device-shell.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00038 | $0.00745 |
| Opus 5.5 | $0.00015 | $0.00298 |
| Sonnet 5.5 | $0.00008 | $0.00149 |
| Haiku 4.5 | $0.00004 | $0.00075 |
Grade A, and why
device-shell scanned grade A with 1 finding 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 2mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sG 'http://127.0.0.1:<port>/exec' --data-urlencode 'cmd=<command>' How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Android Device Shell
Run commands on an Android device from a Linux-based agent environment.
Channel 1: ADB (recommended)
The device is reachable through ADB. Replace <serial> with the actual device serial shown by adb devices.
adb -s <serial> shell '<device shell command>'
Examples:
# identity / basic info
adb -s <serial> shell 'id; getprop ro.product.model; getprop ro.build.version.release'
# list packages
adb -s <serial> shell pm list packages
# launch an app to foreground
adb -s <serial> shell cmd package resolve-activity --brief -a android.intent.action.MAIN -c android.intent.category.LAUNCHER <package>
adb -s <serial> shell am start -n <resolved-component>
# check current foreground app
adb -s <serial> shell dumpsys window | grep -E 'mCurrentFocus|mFocusedApp'
adb -s <serial> shell dumpsys activity activities | grep -E 'ResumedActivity'
Notes:
- ADB shell usually runs as
uid=2000(shell), not root. - Use
am,pm,input,dumpsys,getprop,cmdfor Android-specific operations.
Channel 2: Local Shizuku HTTP bridge
Some Android host apps expose a local HTTP shell bridge. A common pattern is:
curl -sG 'http://127.0.0.1:<port>/exec' --data-urlencode 'cmd=<command>'
Response is JSON:
{"result":"<command output>\n[EXIT=<code>]"}
If it returns a service-not-ready marker such as [SHIZUKU_SERVICE_NOT_READY], the Shizuku UserService has not been bound yet. Restart the host app/service after granting Shizuku permission, then try again.
Channel 3: Termux environment (DSHA)
When the agent runs inside Termux (not proot), destructive commands are wrapped and require user confirmation:
rm,dd,mkfs*,fdisk,reboot,shutdown,halt,poweroff,wipe,pm,sm,settingsare replaced by wrappers in~/dsh-bin/(prepended to PATH by the DSHA launcher).- Running one triggers a
termux-dialogconfirmation popup on the device ("DSHA 安全确认:模型试图执行 [...]"). The command blocks until the user checks "允许" (allow) or cancels. - If rejected, or Termux:API (
termux-apipackage) is not installed, the wrapper exits non-zero. Treat a non-zero exit as "user denied" — do not retry the same destructive command; ask the user or choose a safer path.
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.
- 2mo ago First seen · 81 lines · 38 tokens per session scan A 164319af3762
device-shell is a skill published in the GitHub repository qiannianhuanxiang/DSHA (392 stars, last pushed 2mo ago), licensed MIT. It adds 38 tokens to every session and 745 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
device-shell compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| device-shellqiannianhuanxiang/DSHA | Skill | 392 | 2mo ago | MIT | A | 38 | Claude Code, Codex |
| kotlin-expertpersonamanagmentlayer/pcl | Skill | 44 | 11d ago | Apache-2.0 | A | 55 | Claude Code |
| release-new-versionniki914/zafiro | Skill | 340 | 4d ago | MIT | A | 54 | Claude Code, Codex |
| extract-apkniki914/zafiro | Skill | 340 | 4d ago | MIT | A | 105 | Claude Code, Codex |
| zafiro-qaniki914/zafiro | Skill | 340 | 4d ago | MIT | A | 71 | Claude Code, Codex |
| assistant-add-supportniki914/zafiro | Skill | 340 | 4d ago | MIT | A | 61 | Claude Code, Codex |
| termuxniki914/zafiro | Skill | 340 | 4d ago | MIT | A | 36 | Claude Code, Codex |