TutorialsDaily Routine & Industry Playbook
5 daily actions + 2 maintenance actions + 6 industry scenarios + methodology sources + FAQ + glossary + 3 golden sentences.
5 Daily Actions Quick Reference
| Action | When | Where | Which prompt |
|---|---|---|---|
| Drop materials → ingest | Every time you get new materials | WorkBuddy | "Digest" (Prompt #2) |
| Want to ask → query | Pull knowledge from library | WorkBuddy | Query with sources (#3) |
| Start new → retrieve | Before starting something new | WorkBuddy | Retrieve prompt (#4) |
| Library thin → restock | Topic lacks materials | WorkBuddy | Restock prompt (#5) |
| Library → write | Produce a deliverable | WorkBuddy | Write prompt (#6) |
2 Maintenance Actions
- Reflow: after finishing, write reusable insights back into the library (Prompt #7①)
- Health check: periodically get a read-only health report; report first, fix later (#7②)
One-Sentence Summary
Obsidian views, WorkBuddy works — daily operations are all in WorkBuddy via prompts; Obsidian is the dashboard that makes it readable.
Industry Playbook: How Different People Build Their Library
Remember: the method is one set (build → ingest → query → retrieve → restock → write); what changes is "what materials, what problem."
The industry libraries below are all the same method applied to different scenarios. The authoritative sources for the underlying methodology are in the last section of this lesson.
Key principle: the "materials" in industry libraries must themselves come from authoritative sources — official docs, industry reports, credible media, professional institutions. This is the built-in source constraint of the restock prompt; if the library doesn't have it, say so — label "to be supplemented" rather than using unreliable sources.
Main Case: Career Library (Where Most Start)
- What to build: industry articles + work documents + study materials
- How to use: digest into cards → ask "how does this material relate to my work" → write a work note/review
- What it solves: bookmarked 1000 articles never used, scattered work files, no accumulated learning
- Sample query: Query the current knowledge base: among my saved materials, what reusable methods are there for time management? Distinguish confirmed, inferred, and pending, with clickable source links.
Sales: Client + Playbook Library
- What to build: client follow-up records + winning scripts + product materials
- What to include: client meeting notes, win/loss reviews, product docs, competitor info
- How to use: ask "how to follow up with this client" → retrieve historical scripts → write follow-up plan
- Method basis: PARA's Projects (client projects) + Zettelkasten atomic cards + query retrieval
HR: Policy + Interview Library
- What to build: company policies + interview records + job requirements
- How to use: new hire onboarding Q&A, interview question bank retrieval, policy Q&A
- Sample query: Query the current knowledge base: what are the most common questions new employees ask? Distinguish confidence levels and provide sources.
Teacher / Trainer: Lesson Prep + Material Library
- What to build: courseware + lesson plans + common student questions + teaching materials
- How to use: look up materials for lesson prep, feed past lesson plans to new courseware, auto-retrieve student questions
- Sample query: Query the current knowledge base: what are the most common misconceptions students have about 【topic】? Distinguish confidence levels and provide sources.
Finance: Policy + Report Library
- What to build: tax/financial policies + reports + cases
- How to use: policy Q&A, cross-checking, writing analyses
- Method basis: restock prompt's built-in "only from authoritative sources" (official policy documents, authoritative institution reports) — suited for industries with frequently updated regulations
Content Creator: Material + Review Library
- What to build: materials + revision records + data reviews
- How to use: auto-feed past experience for new topics, write prompt directly produces drafts
- Sample query: I'm about to work on 【a new piece about XX】. First retrieve related materials from the current knowledge base… (retrieve prompt)
Learning Scenarios
| Scenario | What to build | How to use |
|---|---|---|
| Exam prep | Textbooks + notes + mistake bank | Textbooks → note cards → review Q&A |
| Book notes | Chapter-by-chapter note library | Each chapter → cross-topic → comprehensive review |
| Cross-discipline research | Multi-source research library | Papers + articles + reports → synthesis |
Your First Library: Start from One Sentence
Don't know how to begin? Use this template (the three elements of the first prompt):
I usually save __ (what materials), my pain point is __ (specific problem), and I want it to help me ____ (desired outcome).Fill it into the first prompt and send. AI will design, recommend methodology, and interview you — you don't need to figure it all out first; let AI help you clarify. Start from your pain point, build a "small but refined" library (10–30 materials), run the loop, then grow slowly.
Methodology Sources
This method isn't randomly designed — it's backed by validated, authoritative knowledge management theories:
| Methodology | Originator / Source | Core idea | Where used in this course |
|---|---|---|---|
| LLM Wiki | Andrej Karpathy (OpenAI founding member, former Tesla AI Director, Stanford PhD) | AI "compiles" knowledge into a continuously maintained library, not search-on-the-fly | Theoretical foundation: raw/wiki/Schema three-layer structure |
| PARA | Tiago Forte, "Building a Second Brain" | Organize everything with Projects / Areas / Resources / Archives | Structural design reference when building |
| CODE | Tiago Forte, "Building a Second Brain" | Capture → Organize → Distill → Express | Daily actions: ingest / query / write |
| Zettelkasten | Niklas Luhmann (German sociologist) | Write ideas as atomic cards with mutual links; knowledge network grows through links | wiki layer = knowledge cards + bidirectional links |
| Evergreen Notes | Andy Matuschak | Notes should grow continuously and be reusable, not written once | Digested cards = evergreen notes |
| Memex | Vannevar Bush (proposed 1945) | "Personal knowledge machine" — knowledge should be machine-assisted | The intellectual origin of the second brain |
Why "Recommend Methodology Before Building" Is Right
Experts don't "think harder" than novices — they recognize faster "what type of problem is this, which known playbook applies." Their brains hold a vast "pattern library." Finding mature methodology first is thinking like an expert.
Analogy — transfer learning: AI training doesn't start from zero; it loads a "pre-trained base" then fine-tunes, far outperforming training from scratch. Mature methodology is your "pre-trained base."
Applied to AI use: asking AI to "use Pyramid Principle / SWOT / PDCA analysis" produces far better results than "just analyze" — mature methodology constrains output to the right track.
FAQ
| Question | Answer |
|---|---|
| Do I need to code? | No. Everything is chat + prompts |
| Is it safe? | Local files + WorkBuddy only works in your authorized directory |
| Too many materials? | Start with 10–30 "small but refined"; don't stuff everything at once |
| Can I use a different tool? | The method is universal; tools are swappable — the core is "materials → ingest → reflow" |
| Don't know how to express needs? | "I only know the direction is __, ask me questions to help me clarify" |
| Can I move the folder after building? | No — AI remembers the path; moving breaks it |
| How to know if ingest succeeded? | Check if wiki/ has generated card files, if index.md is updated; only a chat summary = not successful |
| How much daily maintenance? | Under 10 minutes a day: digest when you get materials, query when you need, reflow when done |
| Obsidian or WorkBuddy to view? | Daily operations all in WorkBuddy; use Obsidian when you want to see structure or verify sources |
Glossary
Look up any unfamiliar term here.
| Term | Meaning |
|---|---|
| Second Brain | A personal knowledge system that distills learned and saved things into a network, retrievable on demand |
| AI Knowledge Base | Your materials + AI-organized cards + maintenance rules — a system that keeps growing |
| RAG | Retrieval-Augmented Generation: AI searches materials on the fly per query — the "interpreter" path |
| LLM Wiki | AI "compiles" knowledge into the library and maintains it — the "compiler" path, this course's approach |
| raw/ | Source layer: read-only ground truth |
| wiki/ | Knowledge layer: AI-generated cards, summaries, index |
| Schema | Rules layer: config file telling AI how to maintain |
| index | Master index: the library's directory / entry point |
| Ingest | Ingest: digest materials into the knowledge base (corresponds to "Digest") |
| Query | Query: ask the knowledge base questions (corresponds to Q&A) |
| Lint | Health check: inspect the knowledge base for issues |
| ASK / CRAFT / PLAN | WorkBuddy's three modes: answer only / do directly / plan then act |
| Markdown | Plain text formatting (.md files); both Obsidian and WorkBuddy read it |
| Bidirectional link | Mutual links between notes; click to jump |
| Context window | How much background info AI can "remember" at once; knowledge base = AI's external memory |
| PARA | A material organization method: Projects / Areas / Resources / Archives |
| CODE | A workflow: Capture → Organize → Distill → Express |
| Zettelkasten | Luhmann's atomic card + linked note method |
| Evergreen Notes | Notes that grow continuously and are reusable |
| Agent | An AI that works hands-on (WorkBuddy is one) |
Three Golden Sentences
- A knowledge base is not a warehouse — it's a second brain that grows
- AI does the first 4 steps, you do the last 1
- Tools will change; the "materials → ingest → reflow" loop always holds
Tonight's acceptance in one sentence: after sending the first prompt, did AI start interviewing you? If yes = success.
Your second brain starts tonight.
Hands-On: Build Your Library from Zero in 30 min
Steps 0–5: create folder → send first prompt → drop in materials → query with sources → view structure in Obsidian. Each step has acceptance criteria.
Assignment: Build Your Second Brain from Zero
Follow the hands-on walkthrough and run your minimal loop end to end.