simiancraft/simiancraft-skills / skills/ios-simulator-flow-evidence
ios-simulator-flow-evidence
VerifiedCapture proof of a driven iOS-simulator flow: vision-verifiable screenshots, video (mp4), and a manifest tying steps to artifacts, plus an honest list of what a simulator cannot do (deliver a push, use a real camera, place a call, fully observe the network). Produces artifacts to a contract that a separate GitHub-presentation skill consumes; this skill owns extraction, not publishing. Sits beside ios-simulator; project-agnostic. Use for "capture evidence of an iOS flow" or "record the simulator".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.7 KB |
| references/artifact-contract.md | markdown | 2.4 KB |
| references/flow-manifest.md | markdown | 2.9 KB |
| references/screenshots.md | markdown | 2.7 KB |
| references/video.md | markdown | 4.1 KB |
references/flow-manifest.md
View on GitHub---
title: Flow manifest
summary: the manifest.json schema tying each driven step to its artifact and the state it proves, including a device-only list for what the simulator cannot show
status: complete
sources:
- "the manifest is a convention this skill defines; artifact paths reference the formats in artifact-contract.md (PNG, MP4/MOV)"
---
# Flow manifest
Reference for **ios-simulator-flow-evidence**. The manifest is the bundle's index: a small
JSON file that lets a reviewer (or the **prove-work-on-github** skill) understand what
each artifact proves **without re-running the flow**. It lives at the bundle root as
`manifest.json` (see `references/artifact-contract.md`).
## The shape
```json
{
"flow": "sign-in",
"device": { "udid": "<udid>", "name": "<device name>" },
"steps": [
{
"index": 1,
"action": "launch the app",
"artifact": "screenshots/01-launch.png",
"expected": "splash, then the sign-in screen",
"verified": true
},
{
"index": 2,
"action": "tap Sign in",
"artifact": "screenshots/02-signed-in.png",
"expected": "the home screen with the signed-in user",
"verified": true
}
],
"video": "video/flow.mp4",
"deviceOnly": ["push delivery"]
}
```
## The fields
- **`flow`**: the flow name, matching the bundle directory.
- **`device`**: the `udid` and human-readable `name` the flow ran on (discover the booted
UDID per **ios-simulator** `references/lifecycle.md`).
- **`steps[]`**: one entry per driven step:
- **`index`**: the step's order, matching the screenshot's `NN` prefix.
- **`action`**: what was driven (the tap, type, or gesture).
- **`artifact`**: the bundle-relative path to that step's screenshot.
- **`expected`**: the concrete state the step should produce.
- **`verified`**: whether the artifact was confirmed to show `expected` (the
vision-verification from `references/screenshots.md`). A step with `verified: false` is a
recorded gap, not a pass.
- **`video`**: the bundle-relative path to the recording, or omitted if none.
- **`deviceOnly`**: the parts of the flow a simulator cannot prove (push delivery, real
camera or microphone, a phone call, true network conditions; see the honest can't-do list in
`SKILL.md`). Naming them here keeps the bundle from implying coverage it does not have.
## Why it earns its place
The manifest is what turns a folder of PNGs into evidence: it states what each shot was
supposed to show and whether it did, and it is honest about what the simulator left
uncovered. Write it as you capture, not after, so `expected` reflects the intent of the step
rather than a rationalization of the image.
## See also
- `references/artifact-contract.md`: where the manifest sits in the bundle.
- `references/screenshots.md`: the `verified` check the manifest records.
- **ios-simulator** `references/lifecycle.md`: discovering the booted UDID and device name.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |