Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@ios-ship-doctor run preflight on /Users/me/MyApp"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Catch App Store rejections before you hit Submit.
An MCP server that lets your AI assistant — Claude Code, Gemini CLI, Codex, Cursor, Copilot, Windsurf, Zed — diagnose why your iOS app will be rejected — privacy manifests, missing permission strings, unlisted SDKs, leftover test credentials — and then fix them. When Apple does reject you, it pulls the rejection and maps it to a plain-English fix.
Apple's review pipeline silently rejects builds for things you can't see in Xcode:
A required-reason API used in code but not declared in PrivacyInfo.xcprivacy
A permission used with no NS…UsageDescription string (instant crash + reject)
A third-party SDK on Apple's list shipping without its own privacy manifest
A placeholder test credential (like Google's sample AdMob ID) left in Info.plist
Each one costs you a submission cycle — hours or days of round-trips. Ship Doctor checks all of them in seconds and tells you exactly what to change.
Existing App Store MCPs are thin API wrappers that submit your app. Ship Doctor is the one that tells you why it won't pass first.
Related MCP server: App Store Connect MCP Server
No install needed — npx fetches and runs it:
npx -y ios-ship-doctor-mcp preflight /path/to/your/appThen wire it into whichever assistant you use. This is a standard stdio MCP server — it is not Claude-specific and works with any MCP client. Let it print the exact config for yours:
npx -y ios-ship-doctor-mcp config # every client
npx -y ios-ship-doctor-mcp config gemini # just onegit clone https://github.com/menansali/ios-ship-doctor.git
cd ios-ship-doctor
npm install && npm run build
npm link # exposes the `ios-ship-doctor-mcp` commandSupported out of the box: claude, gemini, codex, cursor, vscode, windsurf, zed, generic.
Client | Config file | Root key |
Claude Code |
|
|
Gemini CLI |
|
|
OpenAI Codex CLI |
|
|
Cursor |
|
|
VS Code (Copilot) |
|
|
Windsurf |
|
|
Zed | Zed |
|
Anything else | — |
|
The two footguns the generator handles for you: Codex silently ignores mcp-servers (it must be mcp_servers), and the printed node path avoids version-pinned Homebrew/nvm paths that break on the next upgrade.
Then just ask:
Is
/path/to/my-appready to submit to the App Store?
🩺 Ship Doctor preflight — /Users/you/Desktop/MyApp
VERDICT: NOT READY — 1 issue that commonly causes rejection.
(1 error, 1 warning, 6 passed)
❌ [credential-traps] Google AdMob TEST application ID in Info.plist
Found placeholder/test value "ca-app-pub-3940256099942544".
↳ MyApp/Info.plist
💡 Replace GADApplicationIdentifier with your real ca-app-pub-… ID.
⚠️ [export-compliance] Missing ITSAppUsesNonExemptEncryption in Info.plist
Without this key, App Store Connect prompts about encryption on every submission.
💡 Add <key>ITSAppUsesNonExemptEncryption</key><false/> if you only use HTTPS/TLS.
✅ [privacy-manifest] Privacy manifest consistent with API usage
✅ [usage-descriptions] All permission strings present
✅ [dependencies] SDK privacy manifests OK
✅ [app-icon] App icon asset foundTool | What it catches |
| Runs every check below → single READY / NOT READY verdict |
| Required-reason APIs used but undeclared, invalid reason codes, missing manifest |
| Camera / location / photos / mic / tracking used with no |
| Apple-listed SDKs shipping without a |
| Placeholder / public test credentials in |
| Missing Privacy Policy / Terms of Use (EULA) links on a paywall + in the App Store description (3.1.2) |
| No demo account for App Review (2.1), missing in-app account deletion (5.1.1(v)), social login without Sign in with Apple (4.8) |
| Stripe/PayPal/Paddle for digital content with no StoreKit (3.1.1) |
|
|
| Lorem ipsum, Stripe test keys, |
| Writes a valid |
| Applies the safe fixes automatically; reports the rest as manual follow-up |
preflight also checks export compliance, App Transport Security, app-icon presence, banned APIs (UIWebView), launch screen, version/build sanity, and deployment target. Dependency scanning covers both CocoaPods and Swift Package Manager.
Project layouts understood: classic Info.plist projects, modern projects with GENERATE_INFOPLIST_FILE and no plist file (keys read from INFOPLIST_KEY_* build settings), ios/ subdirectories, and monorepos where the .xcodeproj sits a level or two down.
Ship Doctor reads project files. It cannot see the things most rejections are actually about:
Crashes and incomplete features (2.1) — the most common rejection; needs the app running
Design quality (4.0) and spam/duplicate (4.3) — human judgement
Metadata accuracy (2.3) — whether screenshots match the real app
Privacy label accuracy (5.1.1) — whether declared data collection matches SDK behaviour
Anything a dependency does; source scanning is first-party only
A clean run means "no automated check fired", not "this will be approved."
Tool | What it does |
| Lists your apps (id, name, bundle id) |
| Pulls recent rejections, maps Review Guideline numbers → summaries + fixes |
| The metadata half of preflight: demo credentials actually filled in, review notes, required screenshot sets |
| Explains any Review Guideline number (offline, no key needed) |
Create a key at App Store Connect → Users and Access → Integrations → App Store Connect API, download the .p8, and add env vars to the MCP config:
{
"mcpServers": {
"ios-ship-doctor": {
"command": "ios-ship-doctor-mcp",
"env": {
"ASC_KEY_ID": "XXXXXXXXXX",
"ASC_ISSUER_ID": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
"ASC_PRIVATE_KEY_PATH": "/absolute/path/to/AuthKey_XXXXXXXXXX.p8"
}
}
}
}The .p8 never enters the repo — it stays on your machine and is read at runtime.
If you'd rather not use npm link, point your client at the built file directly (this shape works for Claude Code, Gemini CLI, Cursor, Windsurf and most others):
{
"mcpServers": {
"ios-ship-doctor": {
"command": "node",
"args": ["/absolute/path/to/ios-ship-doctor/dist/index.js"]
}
}
}projectPath in any tool can be your repo root (with an ios/ folder) or the ios/ directory itself.
The same binary runs as a one-shot command that exits non-zero on blocking issues:
npx -y ios-ship-doctor-mcp preflight /path/to/app # human-readable
npx -y ios-ship-doctor-mcp preflight /path/to/app --json # machine-readableCopy examples/preflight-gate.yml into your app repo's .github/workflows/ to block PRs that would fail App Store review.
Required-reason APIs — scans first-party source for Apple's documented API signatures, then diffs against your declared PrivacyInfo.xcprivacy.
App Store Connect auth — an ES256 JWT signed with Node's built-in crypto (no jsonwebtoken dependency), in the JOSE (IEEE-P1363) format Apple requires.
Safe by default — everything is read-only except generate_privacy_manifest with write=true.
Auto-fix tools (patch Info.plist, generate privacy manifest)
More preflight checks: launch screen, version sanity, deployment target
SwiftPM dependency scanning
CLI + CI mode
Entitlements sanity (push, app groups, associated domains)
Privacy nutrition-label cross-check against SDK data collection
Draft reviewer replies from a rejection
Issues and PRs welcome. Adding a check usually means one entry in src/knowledge.ts plus a small function in src/scanner.ts.
MIT © menansali
Check the App Store Connect side of readiness for the next version: whether demo credentials are actually filled in for a login-gated app (Guideline 2.1 — the most common avoidable rejection), whether App Review notes exist, and whether the required iPhone screenshot sets have anything in them. Complements the local preflight, which can only see the binary.
| Name | Required | Description | Default |
|---|---|---|---|
| appId | Yes | App Store Connect app id (from asc_list_apps). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are present, so the description carries the full burden. It explicitly lists what the tool checks (demo credentials, review notes, screenshots), implying a read-only behavior. It does not specify side effects or permissions, but is otherwise transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences with no filler. Each sentence adds essential information: what is checked, why it matters, and how it relates to a sibling tool. Highly efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's simplicity (one parameter, no output schema), the description covers all relevant aspects: purpose, specific checks, related sibling, and a concrete use case (avoiding Guideline 2.1 rejection). No missing information.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% and the description adds value by specifying that the 'appId' comes from 'asc_list_apps', providing context beyond the schema description.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool checks App Store Connect readiness for submission, listing three specific items (demo credentials, review notes, screenshot sets). It uses a specific verb 'Check' and resource, and distinguishes from sibling 'preflight'.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides context on when to use (before submission to avoid rejection) and explicitly names an alternative ('preflight'), clarifying that preflight only checks the binary. It lacks explicit when-not-to-use statements.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List the apps on your App Store Connect account (id, name, bundle id). Requires ASC_KEY_ID, ASC_ISSUER_ID, and ASC_PRIVATE_KEY(_PATH) env vars. Use the returned app id with asc_get_rejections.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden of behavioral disclosure. It mentions the requirement for environment variables (authentication) and hints that the returned id is used elsewhere, but does not discuss side effects, rate limits, pagination, or error handling. For a list operation, basic safety can be inferred, but more detail would improve transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences, front-loaded with the core action and return value, followed by usage guidance (env vars and next step). Every sentence earns its place; there is no redundancy or fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simplicity of the tool (no input parameters, no output schema), the description covers the essential points: what it does, what it returns, prerequisites, and how to use the result. However, it lacks details on whether the list is complete or paginated, and does not describe the exact format of the returned data. For a list tool, this is mostly sufficient but could be slightly more comprehensive.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema has no parameters, so schema coverage is trivially 100%. The description adds value by noting that the tool requires specific environment variables (not parameters) and that no explicit arguments are needed. This is clear and helpful, earning a top score.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'List the apps on your App Store Connect account' with a specific verb and resource. It also specifies the returned fields (id, name, bundle id) and mentions the required environment variables, making the purpose unambiguous and distinct from sibling tools like asc_check_submission.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear context for when to use the tool: to retrieve an app id for use with asc_get_rejections. However, it does not explicitly state when not to use it or mention any alternatives among siblings, which would justify a higher score.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Scan source and Info.plist for content that should never reach review: lorem ipsum, Stripe test keys, YOUR_API_KEY-style template tokens, example.com dead links, and template app names still set as CFBundleDisplayName (Guideline 2.1).
| Name | Required | Description | Default |
|---|---|---|---|
| projectPath | Yes | Absolute path to the iOS project. Can be the repo root (containing an ios/ folder) or the ios/ directory itself. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden. It indicates scanning for specific content types and references Guideline 2.1, suggesting a compliance check. However, it does not disclose whether modifications occur, permissions needed, or the exact nature of the output, leaving some behavioral ambiguity.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, dense sentence that efficiently lists scanned items. It is front-loaded and free of fluff, though a slightly more structured list could improve readability.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the single parameter, no output schema, and no annotations, the description provides a solid overview of functionality. It references a relevant guideline and typical use cases. However, it omits details about the output format or whether results are returned as a list or summary.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
With 100% schema description coverage, the parameter 'projectPath' is well-documented in the schema. The tool description adds no further detail about the parameter, meeting the baseline but not exceeding it.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool scans source and Info.plist for specific placeholder content like lorem ipsum, test keys, and template tokens. The verb 'Scan' is specific and the resource is well-defined, distinguishing it from sibling tools that focus on submissions or app listings.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies use before review to check for unwanted content, which is clear context. However, it does not explicitly state when not to use or directly compare to sibling tools like preflight or explain_guideline, leaving the agent to infer suitable scenarios.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Explain an App Store Review Guideline number (e.g. '5.1.1') in plain language with a typical fix. Works offline — no credentials needed.
| Name | Required | Description | Default |
|---|---|---|---|
| guideline | Yes | Guideline number, e.g. '2.1', '3.1.1', '5.1.1'. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description discloses two important behavioral traits: offline operation and no credential requirement. No side effects or limitations are mentioned, but it is adequate for a read-only explanation tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences achieve maximum clarity with zero waste: the first states the action, the second adds usage constraints. Perfectly front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple one-parameter tool with no output schema, the description covers all necessary aspects: purpose, usage context, and key behavioral traits. Complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema description covers the sole parameter at 100% with examples. The tool description adds no further parameter information beyond what the schema provides, meeting the baseline.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states the tool explains a specific guideline number in plain language with a typical fix, clearly distinguishing it from sibling tools like submission checking or app listing.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description specifies that the tool works offline and requires no credentials, providing clear context for when to use it. However, it does not explicitly exclude alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Generate a valid PrivacyInfo.xcprivacy covering every required-reason API detected in first-party code (plus anything already declared). By default only PREVIEWS the XML; pass write=true to save it to the app target directory. After writing, the file still must be added to the app target in Xcode.
| Name | Required | Description | Default |
|---|---|---|---|
| write | No | If true, write the manifest to <appSourceDir>/PrivacyInfo.xcprivacy. If false, only return the XML for review. | |
| projectPath | Yes | Absolute path to the iOS project. Can be the repo root (containing an ios/ folder) or the ios/ directory itself. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description discloses key behaviors: default preview, write side-effect, and manual Xcode step. It covers scanning first-party code and existing declarations, but doesn't mention overwrite behavior or error cases.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three sentences covering purpose, usage, and post-step with no filler. Front-loaded with key action.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 2-param tool with no output schema, description covers generation purpose, preview vs write, and post-save step. Could mention return value explicitly, but implied.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so baseline is 3. Description restates schema info but doesn't add extra meaning beyond defaults and path flexibility.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool generates a valid PrivacyInfo.xcprivacy covering required-reason APIs. The verb 'generate' and resource 'PrivacyInfo.xcprivacy' are specific. Sibling tools are unrelated, so it distinguishes well.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains default preview behavior and how to save with write=true, plus a post-save step. No explicit when-not-to-use or alternatives, but the tool is unique among siblings and context is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Run ALL App Store readiness checks on an iOS project and return a single prioritized report: privacy manifest vs required-reason API usage, Info.plist usage-description keys, third-party SDK privacy manifests, and placeholder/test-credential traps. Start here — it answers 'is this app ready to submit?'.
| Name | Required | Description | Default |
|---|---|---|---|
| projectPath | Yes | Absolute path to the iOS project. Can be the repo root (containing an ios/ folder) or the ios/ directory itself. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description bears full burden. It describes the tool as running checks and returning a report, without mentioning potential side effects, required permissions, error conditions, or limitations (e.g., handling of invalid project paths). The coverage is adequate but lacks depth.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences with no filler. The first sentence states purpose and scope, the second provides usage guidance. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool has one required parameter and no output schema. The description specifies what the report contains (prioritized, covering specific checks) but does not detail the output format. For a simple tool, this is nearly complete; a small gap in return structure could be improved.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so baseline is 3. The description adds that projectPath can be a repo root or ios/ directory, which is valuable extra guidance beyond what the schema provides ('Absolute path to the iOS project'). This justifies a higher score.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool runs all App Store readiness checks and returns a single prioritized report, listing specific checks (privacy manifest, Info.plist, third-party SDKs, placeholder traps). It positions itself as the starting point, distinguishing from siblings like asc_check_submission or check_placeholder_content.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly says 'Start here' and frames the tool as answering 'is this app ready to submit?', indicating when to use it. It does not explicitly mention when not to use it or provide alternatives, but the context is clear enough for an agent to decide.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0asc_check_submissionasc_list_appscheck_placeholder_contentexplain_guidelinegenerate_privacy_manifestpreflightScored across 6 tools
Each tool serves a distinct, non-overlapping purpose: ASC readiness, app listing, placeholder scanning, guideline explanation, privacy manifest generation, and a comprehensive preflight. No two tools could be confused.
Most tools follow a verb_noun pattern (e.g., check_placeholder_content, generate_privacy_manifest) with a consistent snake_case style. The only outlier is 'preflight' which is a single word, but it is a common term and the overall pattern is clear.
With exactly 6 tools, the server covers the core aspects of iOS app submission readiness without being excessive or lacking. Each tool justifies its presence.
The server covers the full lifecycle of pre-submission checks: placeholder content, privacy manifest generation, Info.plist keys, third-party SDK manifests, ASC credentials and metadata, and a comprehensive preflight that ties it all together. There are no obvious gaps for its stated purpose.
App Store Connect operator for AI agents: icons, TestFlight builds, listings, IAP, rejection fixes.
Run App Store Connect from your IDE: pricing, listings, screenshots, releases, AI visibility.
Compliance & security scan for your app: secrets, exposed files, headers, privacy, AI-disclosure.
AI-agent operations for App Store Connect and Google Play, with approval before live publishing.