🗒️ ChatGPT — Zettelkasten Note Generation for PUMA
Tool: ChatGPT (OpenAI) Phase: Phase 1 — Research Step: 04 — Conceptual Synthesis Methodology: Zettelkasten + PARA integration
Prompt A — Raw Notes to Zettelkasten Conversion
Role: Knowledge organization assistant applying the Zettelkasten method.
Context: I have raw research notes from reading papers on AI agents for project management. I need to convert them into the PUMA Obsidian vault Zettelkasten structure.
Instructions:
1. Classify each raw note into: (a) fleeting note, (b) literature note, (c) permanent note, (d) project note.
2. For permanent notes: write a declarative title (claims a single reusable insight), 3–5 sentence explanation, and link to 2–3 related concepts.
3. For literature notes: include author/year, one-paragraph summary, key quote (verified), and PUMA relevance.
4. For project notes: identify the PUMA phase (F0–F5), one next action, and one risk or dependency.
5. Assign each note to PARA category and Johnny Decimal number.
Format: Structured output per note. Each permanent note should stand alone without context.
Raw notes to process:
[PASTE RAW NOTES HERE]
Prompt B — Synthesis → 10 Permanent Notes
From the research synthesis on AI agents for project management, generate 10 atomic permanent notes for the PUMA Zettelkasten. Requirements for each note:
- Title: declarative sentence claiming one reusable insight (not a question, not a description)
- Body: 3–5 sentences explaining the insight and its evidence base
- Links: 2–3 wikilinks to related Zettelkasten concepts
- Tags: relevant Obsidian tags
- PUMA application: one concrete way this insight informs PUMA's design
Format: One markdown block per note, ready to paste into Obsidian.
Prompt C — GTD Task Extraction from Research
From my research synthesis on PUMA, extract all implied next actions and organize them using GTD:
1. Capture: list every task implied by the synthesis (papers to read, experiments to design, code to write, decisions to make).
2. Clarify: for each task, determine if it is actionable now or a someday/maybe item.
3. Organize: assign each action to: (a) PUMA project folder, (b) Research area, (c) Resource to acquire, (d) Archive.
4. Next Actions: identify the 5 most important next actions for today's session.
Format: GTD capture list → clarified tasks → organized by PARA category → today's 5 next actions.
Prompt D — PARA Organization Prompt
Help me organize the following PUMA research materials into the PARA system:
- Literature PDFs (40 papers)
- Synthesis notes from NotebookLM
- Experimental design decisions
- Dataset download scripts
- Reference bibliography
For each item, classify as: Project (active, has deadline), Area (ongoing maintenance), Resource (reference material), Archive (completed/inactive). Assign a Johnny Decimal number (10-99 range). Explain why each classification makes sense for PUMA's workflow.
PUMA Relevance
ChatGPT’s strength in structured output generation makes it ideal for populating the Zettelkasten. The permanent notes generated here form the knowledge backbone of ZK-Hub-PUMA and connect through bidirectional links to the PUMA project notes.