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 oyi77/1ai-skills --skill phone-harnessgit clone --depth 1 https://github.com/oyi77/1ai-skillsWrote 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/oyi77/1ai-skills/phone-harness)<a href="https://agentmods.dev/skills/oyi77/1ai-skills/phone-harness"><img src="https://agentmods.dev/badge/skills/oyi77/1ai-skills/phone-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.
<a href="https://agentmods.dev/skills/oyi77/1ai-skills/phone-harness"><img src="https://agentmods.dev/badge/skills/oyi77/1ai-skills/phone-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00093 | $0.01132 |
| Opus 5.5 | $0.00037 | $0.00453 |
| Sonnet 5.5 | $0.00019 | $0.00226 |
| Haiku 4.5 | $0.00009 | $0.00113 |
Grade A, and why
phone-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 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.
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.
Phone Harness
When to Use
Trigger phrases:
-
"control my phone"
-
"tap on the phone"
-
"read the phone screen"
-
"install/use phone-harness"
-
One-off phone actions adapters don't cover: OTP/2FA prompts, Settings toggles, account switches, uninstalls, first-run dialogs
-
Debugging a phonefarm device state (what app is foreground, what's on screen)
-
Wiring a new farm workflow before promoting it into a phonefarm adapter
-
Verifying an adapter change against a real device by hand
When NOT to Use
- Anything doable via web/API — leave the phone alone (phonefarm adapters first)
- Recurring production farm workflows → implement in
1ai-phonefarmadapters (adapters/<platform>/mobile_adapter.py), not ad-hoc harness scripts - Tasks needing multi-touch (pinch), camera, or Face ID — unsupported
- iOS flows on this box — no macOS here; Linux workstation = Android-only
Workflow
- Install once (Android focus for this workstation):
git clone https://github.com/ShawnPana/phone-harness ~/.phone-harness cd ~/.phone-harness && pip install -e . # or: read install.md in the repo and follow it phone-harness --doctor # verify adb + device chain - Set Android default (this box has no macOS/iPhone Mirroring):
phone-harness config set platform android - Drive the device (helpers pre-imported):
task: dismiss a first-run dialog on the farm device
step: OCR, tap the right label, verify
open_app("chrome"); wait_stable() tap_ui("Got it") # label from accessibility tree — exact, no misreads
4. Verify after every action — adb reports nothing about outcomes; a tap on
empty space "succeeds". Use `wait_for_app(...)`, `wait_for_text(...)`, or
one `ocr()` check. Batch only steps already watched working.
5. Keep the phone awake for long tasks: `phone-harness android awake --bg`,
then `phone-harness android rest` at the end.
Key helpers: `ocr()`, `ui()`, `find_nodes()`, `tap_ui()`, `tap_text()`,
`screenshot()`, `open_app()`, `back()`, `current_app()`, `list_apps()`,
`swipe()`, `scroll()`, `type_text()`, `press()`.
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 · 119 lines · 93 tokens per session scan A 6e8a62599ee4
phone-harness is a skill published in the GitHub repository oyi77/1ai-skills (15 stars, last pushed today), licensed MIT. It adds 93 tokens to every session and 1,132 once invoked, about $0.0004 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 2026-09-16.
Other skills, from other repositories
extract-apk
A workflow for finding an Android app on a connected device, exporting its APK, and turning it into searchable decompiled source files. An APK is the installable package used by Android apps.
assistant-add-support
A workflow for adapting an Android voice assistant integration to a new app version. It examines the updated app package, checks whether the integration points changed, and creates and verifies a matching configuration.
flutter-error-handling
Implement functional repository error recovery with Either/Failure patterns in Flutter. Use when handling exceptions or dartz Either types in data flows; defer UI-only retry buttons.
rn-agent-observer
Runtime observability for React Native/Expo apps on Android. Use when debugging RN UI, performance (FPS, JS blocking), network, re-renders, or runtime errors on a running app — understand the current screen and UI findings, capture screenshots, UI trees, ref snapshots, per-request network, CDP console/heap/CPU…
flog-inspect
Use when a user asks an agent to inspect Flutter app logs, network traffic, SSE/WebSocket streams, current page state, screenshots, or debugging context through flog instead of copying from the TUI.
android-use
Control Android devices via ADB. Use when user asks to tap, swipe, type, launch apps, take screenshots, or automate Android device interactions.