Dogfood
Exploratory QA on a real iOS or Android build with agent-device—snapshots, screenshots, and runtime-backed issue reports.
Overview
dogfood is an agent skill for the Ship phase that runs exploratory mobile QA with agent-device and evidence-backed issue capture.
Install
npx skills add https://github.com/callstackincubator/agent-device --skill dogfoodWhat is this skill?
- Router skill: agent-device help dogfood defines report shape, evidence commands, and workflow
- Requires agent-device >= 0.14.0; verify with agent-device --version before running
- Loop: open app → snapshot -i + screenshot → explore flows → capture evidence → close
- Allowed tools restricted to agent-device and npx agent-device only—no autonomous global npm upgrades
- Findings must reflect runtime behavior, not static source reads
- Minimum agent-device version 0.14.0 for required help topics
- Allowed-tools scope: Bash(agent-device:*), Bash(npx agent-device:*)
Adoption & trust: 3.3k installs on skills.sh; 2.6k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You are about to ship a mobile app but only have ad-hoc tapping—not a repeatable agent-driven pass that records proof per bug or UX flaw.
Who is it for?
Indie developers with a concrete iOS/Android build installed and a trusted agent-device CLI on PATH.
Skip if: Backend-only APIs, desktop web apps without agent-device mobile targets, or environments where you cannot approve device/bash tooling.
When should I use this skill?
User asks to dogfood, QA, exploratory test, find issues, bug hunt, or test this app on mobile (iOS/Android).
What do I get? / Deliverables
You close a dogfood session with flow coverage, per-issue screenshots/snapshots, and a report shaped by agent-device help dogfood—not guesses from code review.
- Dogfood issue report per agent-device help dogfood
- Per-issue screenshots and interactive snapshots from runtime
Recommended Skills
Journey fit
Dogfooding and bug hunts belong in Ship before you rely on store or production users. Exploratory testing with device evidence is classic pre-release QA, not launch distribution or growth analytics.
How it compares
Runtime mobile exploratory QA via agent-device—not unit tests in CI and not static code review.
Common Questions / FAQ
Who is dogfood for?
Solo mobile builders using Claude Code or similar agents with Callstack agent-device for hands-on QA before release.
When should I use dogfood?
In Ship → testing when asked to dogfood, exploratory test, bug hunt, or QA this app on iOS or Android with real interaction.
Is dogfood safe to install?
It runs device automation via Bash; review the Security Audits panel on this Prism page and only approve trusted agent-device binaries.
SKILL.md
READMESKILL.md - Dogfood
# Dogfood Router for exploratory QA. Private setup before using this skill: ```bash 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: ```bash agent-device help dogfood ``` Loop: open app -> snapshot -i + screenshot -> explore flows -> capture evidence per issue -> close. 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.