Strategic product direction finder for open-source repositories that have reached v1.
Given any working open-source repo, this skill analyzes the product, checks competitors, gathers community signals, and outputs 3-5 prioritized directions for increasing adoption and stars — grounded in data, not guesswork.
| # | Category | Core Question |
|---|---|---|
| 1 | Product Depth | Is the core experience good enough for existing users? |
| 2 | Product Reach | Can we expand to new user groups or adjacent use cases? |
| 3 | Time-to-Value | How fast can a new user feel the value? |
| 4 | Trust & Proof | Why would a stranger believe this is worth their time? |
| 5 | Growth | How do more people discover it, and how do users become promoters? |
- Baseline — Reads the repo structure, README, core features
- External Signals — Checks GitHub Issues, star trends, competitors via WebSearch and
ghCLI - Opportunity Scan — Evaluates each category for gaps
- Rank & Output — Returns 3-5 directions sorted by leverage, each with confidence level and smallest proving move
## 当前判断
- v1 status: ...
- core promise: ...
- main star thesis: ...
## 优先方向
### 1. [Direction]
- type: Product Depth | Product Reach | Time-to-Value | Trust & Proof | Growth
- why this matters: ...
- user disappointment being solved: ...
- why this could increase stars: ...
- confidence: High | Medium | Low
- smallest proving move: ...
## 暂不优先
- [Direction]: why not now- Your repo already works and is public (or nearly)
- You're asking "what should I build next" or "what's missing"
- You want strategic direction, not a task list
- 接下来做什么 / 还缺什么 / 怎么让更多人用
- Brand-new projects that haven't reached v1
- Pure code review, refactoring, or bug fixing
- README-only rewrites
This is a Claude Code skill. Copy SKILL.md to your skills directory:
mkdir -p ~/.claude/skills/proactive-explorer
cp SKILL.md ~/.claude/skills/proactive-explorer/Or clone this repo:
git clone https://github.com/zinan92/proactive-explorer.git ~/.claude/skills/proactive-explorer- Product before presentation — check real capability gaps before suggesting polish
- Data before opinion — competitors and Issues over guesswork
- Directions, not task lists — strategic, not tactical
- Honest — no vanity hacks, no fake traction advice