Athena Notes - Thinking System Foundation

Standardized note types and patterns for the Muse thinking system. This skill defines HOW thoughts are captured, organized, and connected.

Core Philosophy

•Notes are memory - Every note should be findable, linkable, useful later
•Types signal intent - Note type tells readers what to expect
•Links create understanding - Connected notes > isolated notes
•Capture early, refine later - Don't wait for perfect thoughts

Note Types

1. IDEA - Quick Capture

Raw thoughts, sparks, unprocessed inspiration. Low friction, high volume.

When to use:

•Flash of insight
•"What if..." moments
•Something worth remembering but not exploring yet

Filename: YYYY-MM-DD-idea-{slug}.md

markdown

---
type: idea
date: YYYY-MM-DD
tags:
  - idea
  - {topic}
status: captured
sparked_by: "[[source note or conversation]]"  # optional
---

# {One-line description}

{The idea in 1-3 sentences}

## Why This Matters

{Brief context - why capture this now?}

## Questions

- {Question this raises}

## Related

- [[{existing note if any}]]

2. EXPLORATION - Thinking Session

Muse conversations, deep dives, question-driven exploration. The core output of thinking sessions.

When to use:

•Muse session produces insights
•Working through a problem
•Exploring options/approaches

Filename: YYYY-MM-DD-exploration-{topic}.md

markdown

---
type: exploration
date: YYYY-MM-DD
tags:
  - exploration
  - muse
  - {topic}
status: in-progress | complete
session_context: "{what prompted this exploration}"
---

# {Topic} - Exploration

## Context

{What prompted this exploration? What question are we trying to answer?}

## Key Insights

> [!tip] Insight 1: {title}
> {Explanation}

> [!tip] Insight 2: {title}
> {Explanation}

{Or numbered list for simpler insights}

## Open Questions

> [!question] {Most important question}
> {Context or sub-questions}

- {Other question}
- {Other question}

## Threads to Pull

{Future exploration directions}

- [[{Topic for later}]] - {why}
- {Unexplored angle}

## Session Notes

{Optional: Raw notes, quotes, fragments from the session}

3. DECISION - Choice Record

Decisions made, with rationale and consequences. Future-you needs to know WHY.

When to use:

•A choice was made
•Direction was chosen
•Architecture/design decision

Filename: YYYY-MM-DD-decision-{what}.md

markdown

---
type: decision
date: YYYY-MM-DD
tags:
  - decision
  - {domain}
status: decided | pending | superseded
superseded_by: "[[newer decision]]"  # if status: superseded
---

# Decision: {Title}

## Context

{What decision needed to be made? What constraints exist?}

> [!info] Background
> {Additional context if helpful}

## Options Considered

### Option A: {Name}

{Brief description}

| Pros | Cons |
|------|------|
| {pro} | {con} |
| {pro} | {con} |

### Option B: {Name}

{Brief description}

| Pros | Cons |
|------|------|
| {pro} | {con} |
| {pro} | {con} |

## Decision

> [!tip] Chosen: **{Option X}**
> {One-sentence rationale}

{More detailed explanation}

## Consequences

> [!warning] Watch For
> {Risks or things to monitor}

- {Expected outcome}
- {What this enables}
- {What this prevents}

## Related

- [[{Related decision}]]
- [[{Exploration that led here}]]

4. SESSION - Conversation Summary

Full session capture when a thinking conversation was particularly valuable.

When to use:

•End of significant muse session
•Conversation worth preserving in full
•Multiple topics covered

Filename: YYYY-MM-DD-session-{topic}.md

markdown

---
type: session
date: YYYY-MM-DD
tags:
  - session
  - muse
duration: "{approximate length}"
participants: [muse, bryan]
---

# Session: {Topic or Date}

## Summary

{3-5 sentence overview of what was discussed/discovered}

## Topics Covered

1. **{Topic 1}** - {one line}
2. **{Topic 2}** - {one line}
3. **{Topic 3}** - {one line}

## Key Takeaways

- {Takeaway 1}
- {Takeaway 2}
- {Takeaway 3}

## Artifacts Created

- [[{Note 1}]] - {type}
- [[{Note 2}]] - {type}

## Follow-up

- [ ] {Action item if any}
- [[{Topic to explore next}]]

## Raw Notes

{Optional: Unprocessed fragments, quotes, or full transcript}

5. THREAD - Connected Ideas

Meta-note that links related ideas across sessions. Emerges when patterns appear.

When to use:

•Same topic keeps coming up
•Multiple notes are clearly related
•Building toward something bigger

Filename: YYYY-MM-DD-thread-{theme}.md

markdown

---
type: thread
date: YYYY-MM-DD
tags:
  - thread
  - {theme}
status: active | complete | abandoned
---

# Thread: {Theme}

## What This Is

{Brief description of the recurring theme or connected idea}

## Connected Notes

| Date | Note | Contribution |
|------|------|--------------|
| YYYY-MM-DD | [[Note 1]] | {what it adds} |
| YYYY-MM-DD | [[Note 2]] | {what it adds} |
| YYYY-MM-DD | [[Note 3]] | {what it adds} |

## Emerging Pattern

{What's becoming clear as these ideas connect?}

## Open Questions

- {Question the thread raises}

## Next Steps

- {Where this thread might lead}

Capture Triggers

When Muse Should Auto-Capture

Signal	Note Type	Action
"That's interesting" or insight emerges	IDEA	@scribe quick capture
Session explores topic deeply	EXPLORATION	@scribe at natural pause
"I've decided" or choice made	DECISION	@scribe decision record
Session ending, was valuable	SESSION	@scribe session summary
Same topic 3+ times	THREAD	@scribe thread note

Capture Prompts

Muse should use these internal prompts:

code

IDEA CAPTURE:
"An insight just emerged. I'll capture it briefly before continuing."

EXPLORATION CAPTURE:
"We've developed this topic significantly. Let me preserve the key insights."

DECISION CAPTURE:
"A decision was made. I'll record the choice and rationale."

SESSION CAPTURE:
"This session covered valuable ground. I'll create a summary before we close."

THREAD DETECTION:
"This topic has come up multiple times. I should create a thread to connect the dots."

Linking Patterns

Within Notes

Always link to related notes using wikilinks:

markdown

# Good
This builds on [[2026-01-15-exploration-auth]].
As decided in [[2026-01-20-decision-jwt|JWT decision]]...
See [[thread-api-design]] for the bigger picture.

# Bad
This builds on the auth exploration from last week.
As decided earlier...

Backlinks

When creating a new note that relates to existing notes:

•Link TO the existing note from the new note
•Obsidian handles backlinks automatically
•Consider updating the existing note's "Related" section

Tag Hierarchy

code

#idea          → raw thoughts
#exploration   → deep dives
#decision      → choices made
#session       → conversation records
#thread        → connected ideas
#muse          → captured by muse system

#project/{name} → project-specific
#domain/{area}  → knowledge domain

Vault Structure

Recommended Folders

code

.notes/
├── ideas/          → IDEA notes
├── explorations/   → EXPLORATION notes
├── decisions/      → DECISION notes
├── sessions/       → SESSION notes
├── threads/        → THREAD notes
└── inbox/          → Unsorted captures (optional)

Flat vs. Folders

Both work. If using folders:

•Scribe should place notes in appropriate folder by type
•Cross-folder links still work: [[decisions/2026-01-20-jwt]]

If flat:

•All notes in .notes/
•Type is in frontmatter and filename prefix
•Easier to browse chronologically

Default: Flat structure (simpler, works everywhere)

Integration Notes

Loading This Skill

Agents that capture notes should load this skill:

yaml

# In agent frontmatter
skills:
  - athena-notes
  - obsidian

With Obsidian Skill

This skill builds on obsidian skill for:

•Wikilink syntax
•Vault paths
•Frontmatter patterns
•Callout syntax

Load both: skills: [obsidian, athena-notes]

Scribe Integration

Scribe should reference this skill's templates when writing notes. Pass the note type to scribe:

code

@scribe [EXPLORATION] Create a note about our auth discussion:
- Context: We explored authentication options...
- Key insights: JWT preferred because...
- Questions: How to handle refresh tokens?