launch
/launch
Use when the user asks to launch a site or app end to end, or do search, ads and Google Play in one go: "launch the site", "take it live", "запусти сайт и приложение", "сделай SEO, рекламу и гугл плей". Asks once for the facts, runs seo-setup, ads-setup and play-release in order, tracks what waits on Google review, writes TODO.md and ends with a table of each service, its state and next step.
- Author
- workbench-soft ↗
- Repository
- agent-skills ↗
- Licence
- MIT
- Last repository change
- 13.11.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 13.11.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Use when the user asks to launch a site or app end to end, or do search, ads and Google Play in one go: "launch the site", "take it live", "запусти сайт и приложение", "сделай SEO, рекламу и гугл плей". Asks once for the facts, runs seo-setup, ads-setup and play-release in order, tracks what waits on Google review, writes TODO.md and ends with a table of each service, its state and next step.
Triggers from the description
Original author description
Use when the user asks to launch a site or app end to end, or do search, ads and Google Play in one go: "launch the site", "take it live", "запусти сайт и приложение", "сделай SEO, рекламу и гугл плей". Asks once for the facts, runs seo-setup, ads-setup and play-release in order, tracks what waits on Google review, writes TODO.md and ends with a table of each service, its state and next step.
How it works
- 1. Read the project
- 2. Ask once
- 3. Show the plan
- 4. Run the skills
- 5. Waits on Google
- 6. Write TODO.md
- 7. Report back
- Guardrails
SKILL.md preview
--- name: launch description: 'Use when the user asks to launch a site or app end to end, or do search, ads and Google Play in one go: "launch the site", "take it live", "запусти сайт и приложение", "сделай SEO, рекламу и гугл плей". Asks once for the facts, runs seo-setup, ads-setup and play-release in order, tracks what waits on Google review, writes TODO.md and ends with a table of each service, its state and next step.' --- # Launch a site and its Android app The agent MUST run the skills for search, ads and Google Play in sequence, so each one starts with what it needs already there. ## 1. Read the project The repo rules MUST beat this skill. The agent MUST start with `AGENTS.md`, `CLAUDE.md` and the README, then look for: 1. **The site** the agent MUST find its domain, its host, its deploy. 2. **The Android app** what kind it is, or that there is none. A PWA without an app MAY turn into a Trusted Web Activity (TWA). 3. **The work already there** the agent MUST read the README sections on search, ads, Android. Each skill leaves one behind, so a second run resumes where the first left off. The agent SHOULD start with these commands: ```bash ls android pubspec.yaml capacitor.config.* android/twa-manifest.json public/manifest.webmanifest 2>/dev/null git grep -lE 'ca-pub-|ca-app-pub-|google-site-verification|sitemap' -- ':!*.lock' ``` ## 2. Ask once The agent MUST put the questions in a single form, prefilled with what the repo shows: - **Parts** the user MUST tick the ones that apply: search, ads on the site, ads in the app, Google Play. - **Maker** the form MUST ask the name of the person or company for the About page, the store and the ad accounts. - **Contact email** the form MUST ask the one the site, Play and the ad accounts use. - **Audience** the form MUST ask the country and language, which steer content and search engines. - **Accounts** the form MUST ask the Google and Cloudflare accounts for logging in. The agent MUST hand the answers to each skill. A skill MUST NOT repeat a question this step already answered. ## 3. Show the plan The agent MUST post the plan in the chat, then begin with no question: | # | Skill | Part | Needs before it | | --- | -------------- | ------------------------------------------------------------------------- | ----------------------- | | 1 | `seo-setup` | audit, fixes, Cloudflare, Search Console, Bing, content | the site live | | 2 | `ads-setup` | AdSense on the site, up to the review request | step 1 checks passing | | 3 | `play-release` | key, bundle, internal testing | the privacy policy live | | 4 | `ads-setup` | AdMob in a native app, or AdSense hidden in a TWA, then Play declarations | step 3 done | | 5 | `play-release` | a new internal build with the ad code | step 4 code merged | - An unticked part MUST be left out. - A TWA app MUST leave out step 5, unless step 4 changed the start URL. - A native app that has ad code already MUST leave out step 5. ## 4. Run the skills The agent MUST load the skills one after another and follow each, using the step 2 answers. - When the README record and the dashboard both show a step done, the step SHOULD be checked, then skipped. - Every skill MUST commit its own changes, in the repo's commit style. - A skill halted by a wait, such as AdSense review or a Play account ID check, SHOULD NOT hold up the others. The agent MUST note the wait and move to the next step that can run without it. - After two failures, a step SHOULD stop. The agent MUST report the attempts and the message from the dashboard or the command. ## 5. Waits on Google A few steps complete days later. The agent MUST record each one in the target repo's README record, along with its start date: | Wait | Usual length | What to do after | | -------------------------- | ----------------------- | ----------------------------- | | Search Console data | 2 to 3 days | read Pages and Performance | | AdSense review | a few days to 4 weeks | add ad units, turn on serving | | Play developer ID check | a few days | create the app, upload | | Closed test for production | 14 days with 12 testers | apply for production access | | AdMob app review | after the app is public | real ads start serving | ## 6. Write TODO.md In `TODO.md` at the target repo root, each skill keeps its own section. The remaining tasks MUST go under `## launch`: - **Waits** each wait from step 5 MUST be there, with the date to look. - **Next run** the date for the next `launch` run MUST be there. - **One place** a task listed in a skill's section MUST be left out of this one. The section SHOULD read like this: ```markdown ## launch - [ ] Run `launch` again on <date>: AdSense review, Search Console data ``` ## 7. Report back The answer MUST open with a single table, with a row for each service: | Service | State | Link | Next step | | ------------------------ | ------------------------- | -------------- | --------------------------- | | Cloudflare Web Analytics | counting visits | dashboard link | | | Google Search Console | verified, sitemap read | dashboard link | read Pages on <date> | | AdSense | under review since <date> | dashboard link | run `ads-setup` on approval | After it, the answer MUST list: 1. The commits from every skill MUST get a line each. 2. The `TODO.md` tasks MUST get a line each. 3. The date for the next `launch` run MUST be named, with what that run will take up. ## Guardrails - The guardrails of each skill MUST apply here as well. - A release MUST stay on internal testing until the user names a further track. - A change to DNS, the Cloudflare proxy or a payment setting MUST get the user's agreement before it happens.
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · workbench-soft/agent-skills | 100 | MIT |
| devdotfast/whiteboard | 100 | MIT |
| microsoft/vscode | 100 | MIT |
| lilly-logic-labs/lll-skills | 100 | MIT |
| ADWilkinson/astack | 100 | MIT |
| terrylica/cc-skills | 100 | MIT |
| gn01763009/user-thoughts | 100 | MIT |
| Exxzzent/rocket | 100 | MIT |
| ciro-maciel/ccm-stack | 100 | MIT |
Database snapshot: 13.11.2026 · package version 0c41c1b65ad8 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100store-media
Use when the screenshots, video, feature graphic or icon of an app need updating in Google Play, AdMob, AdSense, the site or elsewhere: "upload the screenshots to Google Play", "change the icon everywhere", "залей скрины в гугл плей", "обнови иконку везде". Takes media from… - 100google-ads-evaluate-launch
Read a product launch brief, check current Google Ads state, and write a paid search strategy for the launch to ads/google/strategy/launch-<slug>.md - new keyword opportunities, existing campaigns to update with fresh ad copy, whether a new campaign is warranted, and whether… - 90app-ads
When the user wants to run app install ads, user acquisition (UA), or promote mobile apps. Also use when the user mentions "app ads," "app install ads," "UA," "user acquisition," "Google App Campaigns," "Apple Search Ads," "ASA," "UAC," "App Store ads," "Play Store ads," "CPI,"… - 100seo-setup
Use when the user asks to set up SEO, get a site indexed, register it in Search Console or Bing, add Cloudflare Web Analytics, a sitemap, robots.txt, meta tags or structured data, or prepare for AdSense review: "set up SEO", "настрой SEO", "зарегай сайт в гугле". Finds where DNS… - 90google-ads-strategy
Run Google Search Ads end-to-end: keyword mining, account structure, ad copy, negative keyword architecture, and performance analysis. Covers the full loop from setup to ongoing optimization for B2B SaaS and app makers. Use when the user wants to plan, launch, or optimize Google… - 100google-ads-campaign-launch
Build and launch a Google Search campaign end to end in the current (2026) UI. Use when asked to "set up a Google Ads campaign", "launch Google Search ads", "my Google ads get no impressions", "eligible (limited)", "missing enough relevant keywords", or "what keywords should I…