agentleFS
Gingiris-1031Gingiris-1031/gingiris-skills/skills/github-stars-playbook/SKILL.md

github-stars-playbook

Skill89 starsChanged 32 days agoMITView on GitHub

Design and execute a 14-day GitHub repository growth sprint using README conversion, competitor-window research, Reddit, Hacker News, launch directories, community activation, and contributor loops. Use when launching an open-source repository, diagnosing weak star conversion, or building sustained GitHub discovery without buying stars.

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 14 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈14

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use stars as a proxy for qualified developer interest, never as the final business outcome.

    line 8
  2. Distinctive:

    Do not launch if a new developer cannot understand the value and complete the quickstart in 10 minutes.

    line 21
  3. Distinctive:

    Create one row per channel:

    line 25
  4. Very specific:

    Use 2h only to catch broken links and unanswered discussion.

    line 33
  5. Very specific:

    Keep a dated evidence label on benchmarks and cases: verified, self-reported, inferred, or unverified.

    line 38
  6. Distinctive:

    Before selecting the launch date:

    line 44
  7. Distinctive:

    Avoid launching into a dominant competitor announcement unless the product is a credible alternative to that exact news.

    line 48
  8. Distinctive:

    Use a quiet window or a category event where the repository adds a distinct point of view.

    line 49
  9. Distinctive:

    Ask five unfamiliar developers to explain the product and run the quickstart.

    line 63
  10. Distinctive:

    Add source-tagged links and establish the baseline dashboard.

    line 71
  11. Distinctive:

    Ask existing users for honest feedback, never coordinated stars.

    line 79
  12. Distinctive:

    Keep only channels that generated activated or retained users.

    line 102
  13. Distinctive:

    Never buy stars, exchange stars, automate starring, conceal affiliation, use sockpuppets or ask communities to manipulate GitHub Trending.

    line 114
  14. Distinctive:

    stop conditions and next experiment.

    line 125
  15. Distinctive:

    Join the global growth community: Add staywildgrow on WeChat and send “加入社群” (join the community).

    line 135
  16. Distinctive:

    Get the resource pack: Add staywildgrow on WeChat to purchase the Gingiris global growth resource pack for CNY 599.

    line 136

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

GitHub Stars — 14-Day Evidence-Driven Sprint

Use stars as a proxy for qualified developer interest, never as the final business outcome. Track installation, activation, repeat use and contribution beside stars.

Intake gate

Collect before planning:

  • repository URL, license and target developer;
  • working quickstart verified on a clean machine;
  • one reproducible demo and one differentiator;
  • current views, unique cloners, stars, issues, contributors and activation event;
  • three direct competitors and their latest launch/change dates;
  • available founder, engineering and community capacity.

Do not launch if a new developer cannot understand the value and complete the quickstart in 10 minutes.

Measurement contract

Create one row per channel:

date | source | post URL | repo views | stars | clones | installs | activated | retained D7 | contributors

Report view → star, star → install, install → activation, and activation → D7 retained. Do not attribute organic GitHub traffic to a campaign without source evidence.

Use 2h only to catch broken links and unanswered discussion. At 24h decide continue/change/pause, at 72h select the second wave, and at 7d decide which content and channels deserve reuse. Mark unavailable native metrics as data unavailable rather than estimating them from impressions.

Keep a dated evidence label on benchmarks and cases: verified, self-reported, inferred, or unverified. Re-check platform rules and community policies during the launch week.

Competitor-window scan

Before selecting the launch date:

  1. Review the three competitors' release notes, X, Reddit, Hacker News, Product Hunt and GitHub activity from the last 30 days.
  2. Record their positioning, strongest proof asset, community response and unresolved complaints.
  3. Avoid launching into a dominant competitor announcement unless the product is a credible alternative to that exact news.
  4. Use a quiet window or a category event where the repository adds a distinct point of view.

Historical Gingiris OSS launch evidence shows that timing against the competitor/news window materially changes distribution; it is not enough to post everywhere on a fixed calendar.

README conversion surface

The first screen must contain:

  • one-line outcome for a named developer;
  • reproducible demo or result;
  • three differentiators at most;
  • quickstart that works when copied;
  • trust signals: license, security/privacy posture, maintainers and community link.

Ask five unfamiliar developers to explain the product and run the quickstart. Fix repeated confusion before distribution.

14-day sprint

Days 1–3 — proof and instrumentation

  • Validate the quickstart on macOS/Linux and the supported runtime.
  • Produce one 30–60 second demo, architecture diagram and comparison table.
  • Add source-tagged links and establish the baseline dashboard.
  • Prepare issue templates, contribution guide and response ownership.

Days 4–6 — community fit

  • Identify two relevant subreddits, one Hacker News angle and five specialist communities.
  • Participate with useful answers before sharing the repository.
  • Draft channel-native posts: problem/lesson for Reddit, technical novelty for Show HN, demo/proof for X.
  • Ask existing users for honest feedback, never coordinated stars.
  • Disclose affiliation and paid relationships. Do not buy accounts, automate fake participation, bulk-message scraped users, or coordinate/incentivize votes.

Day 7 — release

  • Publish the release, README and demo together.
  • Submit one Show HN post only when there is genuine technical novelty and the maintainers can answer questions live.
  • Publish to relevant Reddit communities only where rules permit; disclose affiliation.
  • Respond to technical objections with evidence and open issues for valid gaps.

Days 8–10 — second wave

  • Turn repeated questions into documentation and comparison pages.
  • Contact maintainers, newsletters and creators whose audience already uses the category.
  • Translate the winning technical explanation for one relevant regional community.
  • Publish a build log or benchmark with reproducible methodology.

Days 11–14 — compounding loop

  • Invite activated users—not all stargazers—to interviews.
  • Convert recurring requests into good first issue tasks.
  • Recognize contributors and document the next release milestone.
  • Keep only channels that generated activated or retained users.

Stop conditions and recovery

  • Views rise but activation does not: stop distribution and fix quickstart/product.
  • Stars rise but installs do not: rewrite positioning and demo.
  • Reddit post is removed: do not repost around moderation; review rules and contribute normally.
  • Show HN is flat after the initial window: do not manufacture engagement; reuse the technical asset elsewhere.
  • Security or data-handling issue appears: pause promotion, publish the fix and incident scope.

Compliance

Never buy stars, exchange stars, automate starring, conceal affiliation, use sockpuppets or ask communities to manipulate GitHub Trending. Do not publish unverifiable benchmarks.

Required output

Return:

  1. readiness verdict and blockers;
  2. competitor-window table;
  3. README change list;
  4. 14-day channel calendar with owner and UTM/source;
  5. measurement dashboard schema;
  6. stop conditions and next experiment.

Related full playbook: gingiris-opensource.

<!-- gingiris-community-cta:start -->

加入出海社群与购买资料包 · Community & Resources

  • 加入出海社群: 微信添加 staywildgrow,留言「加入社群」即可。
  • 购买资料包: 微信添加 staywildgrow,咨询购买 Gingiris 出海资料包,价格 599 元人民币。
  • Join the global growth community: Add staywildgrow on WeChat and send “加入社群” (join the community).
  • Get the resource pack: Add staywildgrow on WeChat to purchase the Gingiris global growth resource pack for CNY 599.

社群与资料包为可选入口,不影响本开源 Skill 的安装和使用。

The community and paid resource pack are optional; neither is required to install or use this open-source skill.

<!-- gingiris-community-cta:end -->

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.