ShipASO

io.github.airowe由 GitHub 账号 airowe 发布v0.1.0更新于 Oct 24, 2026

Audit any App Store or Google Play listing: measured ranks, keyword gaps, draft copy. Read-only.

发布方自己的服务:ship.page

已验证Streamable HTTP可网页运行数据与分析商业与电商

概览

AI 生成的概览

ShipASO 是一个托管的只读应用商店优化服务器,让助手审计 App Store 与 Google Play 列表、查看排名与关键词缺口,并起草优化文案。

可以做什么
预览与审计列表:对任意 App Store 应用和 Google Play 列表提供只读 ASO 快照与按优先级排序的审计,并可通过 Play Developer API 审计自己的 Play 应用,另有截图评分与覆盖率检查。研究关键词与排名:测量目标关键词的自然排名,找出相对竞品的关键词缺口,并给出按可赢性排序的机会。跟踪竞品与历史:对比竞品列表与上次状态的变化,将你的排名历史与竞品正面比较,并为已连接应用给出按投资回报排序的本地化建议。起草文案:返回带校验的优化列表草稿(名称、副标题、关键词、描述),不保存也不推送。
适用场景
当你希望助手评估应用商店列表、测量真实关键词排名、与竞品比较,或在不触碰线上商店的情况下起草改进后的元数据时使用。公开工具无需账号即可快速查看任意 App Store 应用。
运行要求
通过 streamable HTTP 连接的远程 MCP 端点,无需本地运行时或安装包。公开层无需密钥。带密钥的工具需要免费的 ShipASO 密钥,以 Authorization 请求头传递,格式为 Bearer 加密钥。自有 Play 审计需要配置 Play 服务账号(GOOGLE_PLAY_SERVICE_ACCOUNT)。
安装前请注意
Authorization 请求头携带 ShipASO API 密钥,需妥善保管。自有 Play 审计需要 Google Play 服务账号这一敏感凭据。所列工具均为只读或仅草稿,不会推送到商店;但 README 提到另有托管付费层,含按月订阅,发布需人工审批。

安装

在 SourceWeft 中

  1. 打开 控制台中的 ShipASO,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.airowe%2Fshipaso/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。

{
  "mcpServers": {
    "shipaso": {
      "type": "http",
      "url": "https://api.shipaso.com/mcp"
    }
  }
}

README

ShipASO

The reason → execute ASO loop for App Store and Google Play.

The product is ShipASO; the codebase, this repo, and the Cloudflare services are named store-ops.

Every other AI ASO tool does one half. Advisory tools (and most "AI ASO agents") reason about your listing then stop at the copy-paste boundary — and they depend on a paid data API. Deploy CLIs (Fastlane, raw APIs) execute but have no ASO brain. ShipASO does both, on both stores, with no paid data dependency.

The open lane: Google Play. There is no public tool that operationally optimizes a Google Play Console listing end-to-end. ShipASO does iOS and Android.

Install — any coding agent

The loop runs in your editor, on your machine, with your credentials. The 33 skills are plain Markdown runbooks over plain CLI tools, and the engine is standard-library Python — so nothing here is tied to one vendor.

Any agent (Codex, Cursor, Copilot, Gemini, Amp, Aider…):

bash
git clone https://github.com/airowe/store-ops && cd store-ops

Then point your agent at AGENTS.md — most read it automatically. That file is the front door: the skills index, the credential rules, and how to run the engine. Cursor, Copilot, and Gemini also get a pointer file that redirects there.

Claude Code additionally has a plugin installer:

/plugin marketplace add airowe/store-ops/plugin install store-ops@store-ops

Either way, audit any listing — read-only, no credentials, no keys:

bash
python3 lib/aso_rank_check.py --bundle com.burbn.instagram "photo,stories,reels"

You get a per-field ASO score of your live listing and the exact next skill to run for each gap. → Full walkthrough + what you'll see: https://shipaso.com/install

Zero install: point your agent at the hosted MCP

No clone, no key, no account:

claude mcp add shipaso --transport http https://api.shipaso.com/mcp

Your agent immediately gets preview_app — a real, measured teaser of any App Store app (audit grade, lead organic rank, how many keywords crack the top 10) — and proof, the aggregate rank-win record. The other ten tools (listing audit, keyword gaps, rank check, screenshot scoring, competitor watch, war room, localization gaps, draft copy, and the two Google Play audits) unlock with a free shipaso_ key from app.shipaso.com → Settings → Agent access; re-add the server with --header "Authorization: Bearer <key>". MCP is vendor-neutral, so Cursor and any other MCP client work the same way. Every tool is read or draft; nothing pushes to a live store. Details: skills/shipaso-mcp/SKILL.md.

What's in this repo

ShipASO comes in two forms — run it yourself, or let the hosted agent run it for you:

PathWhat
skills/ + lib/The free OSS agent plugin — 33 skills + the engine (Python, 200 tests). Runs in any coding agent; Claude Code also has a one-command installer.
cloud/The hosted autonomous agent — a Cloudflare app (Workers + D1 + Cron + Pages) that runs the loop on a schedule and surfaces decisions for approval. Engine ported to TypeScript (2,160 tests). See cloud/README.md + cloud/DEPLOY.md.
commercial/, docs/The offer, the launch posts, the landing page.

The plugin is the funnel (discover the agent by running it); the hosted app is the product (it keeps working while you build). Same engine, two surfaces.

Run it yourself, or let it run itself:

  • Free, in your editor — install the plugin and run the loop yourself in whichever agent you already use. Your credentials, your machine, no cost. Start here ↓ (Install).
  • Hosted + autonomous — connect an app by bundle id and an agent runs the loop on a weekly schedule: re-checking ranks, watching competitors, drafting the next optimization, surfacing each decision for you to approve. It's live: https://app.shipaso.com

What it does

audit  →  research keywords  →  optimize to char limits  →  push  →  verify rank(asc/gplay) (no paid API)        (ready copy + commands)   (you approve) (free, over time)

The full loop, not half of it. The data tools (Astro, AppTweak, Sensor Tower) stop at "here's what to do" and leave you in the App Store Connect form. ShipASO picks the keywords, writes the metadata, hands you the push, and then verifies the rank actually moved — the only piece that closes the loop and the one nobody else ships.

/store-ops myapp                     # full chain, both stores/store-ops myapp --store playstore   # Google Play only/store-ops myapp --audit-only        # just score the live listing

Try it in 30 seconds (no credentials, no setup)

The verify/watch half runs on the free public iTunes API — try it on any live app right now:

bash
# organic App Store rank for some keywords (any app's bundle id)python3 lib/aso_rank_check.py \    --bundle com.burbn.instagram "photo,stories,reels"
# score an app's screenshots against ASO best practicepython3 lib/aso_screenshot_score.py \    --app instagram --bundle com.burbn.instagram
# resolve a non-US market's keyword-volume + rank constantspython3 lib/aso_locale.py --locale de-DE

Run the test suites (standard-library only — no network, no keys):

bash
python3 lib/run_tests.py

Skills

Reasoning (the IP):

SkillDoes
aso-auditScore a live iOS/Android listing field-by-field vs ASO best practice. Read-only.
issue-verifyCheck open issues against the source before working them — catches stale "this is missing" claims.
aso-teardownWrite a publishable long-form teardown of any app from public measured data. Never estimates revenue.
aso-keyword-researchRank keywords from store autocomplete + competitor scrape + synonyms. No paid data API.
aso-metadata-optimizationGenerate final copy at exact char limits + emit the push commands. Never auto-ships.
aso-rank-checkRead your organic App Store rank per keyword and log it over time — did the change land? Free public iTunes API, no key.
store-opsRouter for the full chain.

More reasoning — audit, risk, and research:

SkillDoes
aso-contextScaffold the per-app context.md every other ASO skill reads. Start here.
aso-review-mineMine your own reviews for the words real users use to describe the app.
aso-offstore-mineMine "best <category> apps" articles + YouTube for off-store discovery language.
aso-localize-researchKeyword + rank research per market (de-DE, ja-JP, es-MX…) instead of US-only.
aso-competitor-watchTrack competitors' listing changes over time — name, subtitle, price, rating.
aso-screenshot-scoreScore a live screenshot set — count, device/aspect coverage, first-two story.
aso-review-riskFlag copy that gets apps rejected — unverifiable claims, "#1", competitor names.
aso-rejection-assistantTurn a rejection into a plan against the guideline the reviewer actually cited.
aso-ppo-treatmentDesign a free screenshot A/B test (Product Page Optimization).
aso-rank-monitorRun a keyword set on a cadence so you get deltas, not one snapshot.

Execution — App Store Connect (via the asc CLI):

SkillDoes
asc-metadata-write-laneCreate the editable version, attach the build, push the copy — stops before submit.
asc-screenshot-write-lanePut an approved run's screenshot strip on the editable version through ShipASO — ids resolved, per-shot ledger, stops before submit.
asc-metadata-syncSync + validate metadata; migrate the legacy metadata format.
asc-localize-metadataTranslate and sync metadata across locales.
asc-submission-healthPreflight a submission, submit a build, monitor review status.
asc-id-resolverResolve app / build / version / group IDs from human-friendly names.
asc-ppp-pricingPer-territory pricing by purchasing power.
asc-shots-pipelineiOS screenshot capture via xcodebuild / simctl / AXe.
aso-goldie-configEmit a goldie.config.ts from a run — audit-ordered scenes with headlines from the proposed copy, rendered on your own Mac.

Execution — Google Play (via the gplay CLI):

SkillDoes
gplay-metadata-syncListing sync, including Fastlane format.
gplay-review-managementReview monitoring, filtering, and responses.
gplay-rollout-managementStaged rollout orchestration and monitoring.
gplay-vitals-monitoringCrashes, ANRs, performance, errors.
gplay-screenshot-automationAndroid capture across devices/locales via adb + Espresso.
gplay-ppp-pricingRegion-specific pricing.

Hosted:

SkillDoes
shipaso-mcpPoint your agent at the hosted ShipASO MCP server and drive the loop over it.

Every asc-* / gplay-* skill that writes asks which API key to use before it does — asc resolves credentials from a default profile, and with more than one registered a push through the wrong one succeeds against the wrong account.

Principles

  • No paid data dependency — reasons over free + owned data (store autocomplete, public listings, your own asc/gplay exports). Every credentialed source is an optional accelerator with a free fallback, never a gate:
    • volume — bring your own Apple Search Ads / Google Keyword Planner keys for real search-popularity numbers; without them, an honest autocomplete-rank proxy.
    • scrape — bring a context.dev key for clean competitor-listing + brand-data scraping; without it, WebFetch / Crawl4AI. You bring your own keys (it's your data, ToS-clean); the plugin never resells data.
  • Both stores, official APIs.
  • Nothing ships without your approval — reasoning writes copy + prints commands; you run the push.
  • App-agnostic — --app <slug> + a context.md. No hardcoded apps, no secrets in the plugin.

Requirements

  • To execute: the asc and gplay CLIs installed + authed with your own developer credentials.
  • The reasoning skills work standalone (they emit commands you can run later).

Install

/plugin marketplace add airowe/store-ops/plugin install store-ops@store-ops

Open-core — what's free vs. hosted

The plugin is free and MIT-licensed, forever. Everything above — the full audit → research → optimize → push → verify loop, both stores, all 33 skills — runs locally with your own credentials at no cost. Use it, fork it, ship apps with it. That's the whole product for anyone who's comfortable in a terminal.

The line for the hosted tier is convenience, never capability. The free plugin is complete; the paid tier sells you out of the setup pain and the remembering. The hosted agent is live now at https://app.shipaso.com — connect an app and watch the loop run on a schedule.

Free (this plugin)Hosted agent
The 33 skills + full loop✅ all of it✅ same engine
Real volume dataBYO Apple/Google keysguided setup
Rank trackingrun aso-rank-check by handscheduled weekly + history + delta alerts
Standing autonomy (weekly cron)❌ you re-run it✅ Indie / Startup / Scale
Appsunlimited (it's local)1 (Free) · 3 (Indie) · 10 (Startup) · 50 (Scale)
Approval gateyou run the pushenforced in code — commands withheld until you approve

The tiers:

TierPriceWhat you get
Free$0Run the agent yourself in any coding agent. The whole loop, your machine. 1 connected app, manual runs only.
Indie$6.99/monthThe weekly autonomous sweep, up to 3 apps. It proposes, you approve, it ships.
Startup$19.99/monthThe weekly sweep across up to 10 apps, with rank history and competitor watch.
Scale$64.99/monthUp to 50 apps plus the portfolio roll-up. For agencies and multi-app devs.

We will never resell Apple/Google data through a shared account (that breaks their ToS) — the hosted tier is BYO-credentials too, and we never hold your store credentials; the push is a generated-commands handoff. What you pay for is the guided onboarding, the scheduler, and the history. The differentiator isn't the data — everyone has Apple's popularity numbers. It's that ShipASO ships the metadata and proves the rank moved.

→ Try the hosted agent: https://app.shipaso.com

License

MIT — see LICENSE.

来源:README.md,提交 7cd1b7e

工具

12

preview_app

Read-only ASO snapshot of any App Store app (grade, lead rank, top-10 count, sample). No account or store push — the optimized copy + push commands stay behind the human-approved loop. Works without a key.

输入参数querycountrybundleId

audit_app

Read-only listing audit: prioritized findings (severity + impact) plus a summary, derived from the live public listing. Read-only — never edits or pushes.

输入参数querycountrybundleId

audit_play_app

Read-only GOOGLE PLAY listing audit: screenshot grade, the 30/80/4000 title/short/long-description budget, target-term coverage in the long description (Play's keyword surface), a keyword-stuffing guard, prioritized findings + summary, and capability locks for surfaces a public read can't see. Takes a Play package id (com.foo.bar) or a play.google.com URL — NOT a free-text name (Play has no public name search). Play has NO keyword field. Reads public Play data only; never edits or pushes.

输入参数brandquerycountrytargetspackageName

audit_play_app_owner

Read-only audit of YOUR OWN Google Play app via the official Play Developer API — full fidelity including the short description (which the public page can't show), with NO capability locks. Requires a configured Play service account (GOOGLE_PLAY_SERVICE_ACCOUNT). Reads only — it opens and DISCARDS a Play 'edit' and NEVER commits, so it can't publish. Owner-only; for a competitor use audit_play_app.

输入参数brandtargetslanguagepackageName必填

keyword_gaps

Read-only keyword opportunities: terms tracked competitors VISIBLY use that this app doesn't target and doesn't rank top-50 for, sorted by winnability. Inferred from competitor name/subtitle only — never their ranking algorithm.

输入参数querycountrybundleId

rank_check

Read-only organic rank check across the app's target keywords, plus winnability-ranked opportunities computed from those measured positions. Positions are live-measured, never fabricated.

输入参数querycountrybundleId

screenshot_coverage

Read-only screenshot scoring for the app's live store gallery (count, score, grade, levers to improve). Read-only — never uploads or pushes assets.

输入参数querycountrybundleId

competitor_watch

Read-only competitor diff: resolves the named competitors and reports their current listings, what changed vs. the last seen state, and a digest line. Read-only metadata comparison; no ranking-algorithm data is invented.

输入参数querycountrybundleIdcompetitors必填

war_room

Read-only head-to-head rank war room for one of YOUR connected apps: your tracked-keyword rank history vs. live competitor positions on those same keywords. Reads stored data + live competitor ranks; writes nothing.

输入参数appId必填competitors

localization_gaps

Read-only localization expansion recommendations (ROI-sorted locales to add) for one of YOUR connected apps, from its most recent App Store Connect run. A static, bundled heuristic over the locales already read — no live install data is fabricated. Empty until you've run an ASC-connected pass.

输入参数appId必填

proof

Read-only, anonymized aggregate proof across all tracked apps (real rank-win numbers — no app names, no user data). This is the 'prove the rank moved' surface that closes the prepare → approve → push → prove loop. Works without a key.

无需输入参数。

propose_copy

Returns a DRAFT optimized listing (name/subtitle/keywords/description) with validation for any app. THIS IS A DRAFT ONLY: it does NOT persist a proposal and does NOT push to the App Store. Publishing stays a human-approved action in ShipASO — an agent can draft and read; only a human approves and pushes.

输入参数querycountrybundleId

版本历史

1
  1. v0.1.0最新Sep 16, 2026

评分

已满足适用检查项的 75%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分

    请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。

  • 已声明认证方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分

其他 Ship 集成