dogfood
- Use it when:
- You want an app dogfooded, bug-hunted or QA'd on an iPhone, Android phone, simulator or emulator.
- It does:
- Opens the app, maps its navigation, works through main flows and edge states, and records evidence for each issue as it finds it.
- You get:
- ./dogfood-output/report.md with severity counts and, per finding, repro commands, expected and actual behavior, and screenshots or video.
- Needs:
- agent-device 0.14.0 or later, on your PATH or through an exact-version npm command you approve; it won't install or upgrade it on its own.
- Judges:
- Only what the running app does, never its source code.
- Unlike agent-device:
- That one is the repo's skill for driving an app; this one runs a whole QA pass and writes the report.
TL;DR
A short router: the SKILL.md checks the agent-device version, then reads the CLI's own guide with agent-device help dogfood, so the workflow matches the installed CLI. That guide has the agent find the target app, sign in or ask you for a code if needed, capture a baseline snapshot and screenshot, map the top-level navigation, then work through main flows and edge states: forms, empty, error and loading states, offline and retry, permissions, settings, accessibility labels, orientation and keyboard, and obvious stalls. Each issue gets its evidence on the spot, step screenshots and usually a video for interactive bugs, one screenshot for static ones, and a severity from critical to low. React Native error overlays count as findings or blockers, and Expo Go and dev-client URLs are handled. It aims for 5 to 10 well-evidenced issues, reports the ground covered and the remaining risk if it finds none, keeps every artifact, and closes the session, shutting down any simulator it booted unless you want it kept. Its allowed tools are limited to agent-device commands.
Summarizes the skill at 463fd7b.
npx skills add https://github.com/callstack/agent-device --skill dogfood skills.sh pageSource on GitHubAll skills from callstack/agent-device
- Total Downloads
- 12,400 on skills.sh, Nov 9, 2026
- Repo stars
- 5,303 callstack/agent-device
- First seen
- Mar 2, 2026 on skills.sh
Weekly values
| Week to | Installs |
|---|---|
| Sep 21 | 595 |
| Sep 28 | 648 |
| Oct 5 | 690 |
| Oct 12 | 681 |
| Oct 19 | 531 |
| Oct 26 | 521 |
| Nov 2 | 543 |
| Nov 9 | 551 |
| Category |
Find root causes, review code and test what you built. Drive desktop apps, phones and simulators: tap, type, read the screen and test native apps. Everything for building software: planning, coding, debugging, design, frameworks, cloud, agents and skills. |
|---|---|
| Type |
A multi-step process the agent follows: interview, debug, run a TDD loop, review. |
| Effect |
Its own steps create or change files, git state or local tools. |
| Needs |
A specific command-line tool or desktop app must be installed, such as gh, agent-browser, LibreOffice, Obsidian or a vendor CLI. |
| Publisher |
A verified publisher on skills.sh, or the company or maintainer behind the product the skill operates. |
| Topics | |
| Stack |
SKILL.md
From callstack/agent-device at 463fd7b. Checked daily, last changed Nov 6, 2026.
Frontmatter
name: dogfood
description: Systematically explore and test a mobile app on iOS/Android with agent-device to find bugs, UX issues, and other problems. Use when asked to dogfood, QA, exploratory test, find issues, bug hunt, or test this app on mobile.
allowed-tools: Bash(agent-device:*), Bash(npx agent-device:*)Dogfood
Router for exploratory QA. Private setup before using this skill:
agent-device --version
If that fails, stop and tell the user to expose a trusted agent-device binary on PATH or approve an exact-version npm command. This skill intentionally keeps allowed tools restricted to agent-device and npx agent-device.
Require agent-device >= 0.14.0; older CLIs lack these help topics. If older, stop and tell the user to upgrade the trusted install or approve an exact-version npm command. Do not run npm install -g agent-device@latest or npx -y agent-device@latest autonomously, and do not include version/upgrade commands in final plans.
Read current CLI guidance:
agent-device help dogfood
Loop: open app -> snapshot -i + screenshot -> explore flows -> capture evidence per issue -> close (close --shutdown if your session booted a local simulator or emulator, unless the user asks to keep it running).
Target app is required; infer platform or ask. Findings must come from runtime behavior, not source reads. Let help dogfood provide exact report shape, evidence commands, and current workflow guidance.