English | 中文
An OpenClaw Skill that turns your 🦞 into an inspiration capture assistant for Obsidian.
Instead of just summarizing content, it helps you articulate: why it's worth remembering, how to classify it, and what to do with it next.
Typical AI-powered note capture:
Drop content → AI summarizes → Save to vault
Sounds great, but in practice the AI tags and categorizes just fine — it just can't capture why you cared about this in the first place. You end up with notes full of technically correct but personally meaningless summaries.
This Skill changes the flow to:
Drop content → Extract → Check if "why" is missing → Ask follow-up → Archive → Add value judgment
The key difference: organizing is a conversation, not compression.
Every captured inspiration automatically gets four sections:
Why did this resonate with you? How does it relate to your current projects, content direction, or decision-making framework?
This is the most critical layer. AI handles structure, but only you know why something matters to you. If you haven't articulated it, the Skill asks — instead of making something up.
| Category | Meaning |
|---|---|
| Action Trigger | Makes you want to act immediately, convertible to short-term action |
| Slow Ferment | Not immediately useful, but shapes long-term judgment and direction |
| Reference Library | Samples that can be cited in future articles, products, or talks |
| Content Seed | Has expressive potential, but isn't a complete piece yet |
- Build Something: Can this become a feature, product, or workflow?
- Create Content: Can this become an article, topic, or long-term perspective?
Concrete todos. Not "look into it sometime" — what's the actual next step.
When you drop an inspiration but haven't explained why it matters, the Skill proactively asks:
- Why do you want to save this?
- Is this more about building a product, creating content, improving a workflow, or developing a mental framework?
- Which of your current projects or themes does this relate to?
- What's the single most important point you want to preserve?
- Is this an action trigger or slow ferment material?
The follow-up questions aren't there to annoy you — they force you to think clearly. The act of writing a note about why is itself the thinking process.
The Skill automatically enters inspiration capture mode when you share:
- A link / screenshot / post
- A quick thought / product judgment
- A rough content idea
- An "oh that's interesting" moment
When triggered, it always does two things: discuss + archive. Discussion without archiving = incomplete.
A typical AI summary:
## Key Points
- The article discusses trends in AI agents
- Multiple use cases are mentioned
- The author believes the future looks promising
After this Skill processes it:
## 📋 Key Points
Agents shouldn't aim to "automate everything" — they should
"provide structure where humans are uncertain."
The author uses an analogy: an agent is scaffolding, not the building.
## ❓ Why Worth Recording
The "scaffolding vs building" metaphor matches my experience designing
OpenClaw Skills. A Skill shouldn't think for the user — it should help
the user turn vague ideas into structured output.
## 🧩 Classification
Slow Ferment — won't act on it now, but it'll shape how I design Skills
## 💡 Value to Me
- Build: When designing a new Skill, ask "is this scaffolding or the building?"
- Content: Could write a piece on "why my AI assistant shouldn't think for me"
## ⚡ Next Action
- [ ] Add the "scaffolding metaphor" to my Skill design principles
The difference: the first captures what AI thinks is important. The second captures what I think is important, with AI helping me say it clearly.
Place SKILL.md in your OpenClaw workspace skills directory:
~/.openclaw/workspace/skills/obsidian-brain-capture/SKILL.md
Then update the paths in SKILL.md to match your environment:
vault_path: /path/to/your/vault/
inbox_path: OpenClaw/Inbox/Three core insights behind this Skill, learned from community discussions:
- AI handles structure, you handle "why" — writing the note is the thinking
- Organizing is a conversation — let AI ask you, instead of guessing silently
- Teach by example first — different types of inspiration need different treatment
These became the follow-up rules + classification framework + four-section template.
MIT