transforms raw transcriptions into polished obsidian notes with zero information loss. use for meetings, lectures, or interviews requiring cognitive reorganization, obsidian callouts, and structured formatting.
Transform raw, machine-generated transcriptions into polished, cognitively-ordered Obsidian notes that are both readable and complete.
Convert poorly transcribed audio/video content (workshops, lectures, meetings, interviews) into well-structured, publication-ready documents while preserving 100% of the original information.
[!danger] Critical Requirement Never omit, summarize, or compress information from the original. Every detail, example, tangent, question, and answer must be preserved. The output should contain MORE structure, not LESS content.
Transform stream-of-consciousness speech into logical document sections:
| Speech Pattern | Transforms To |
|---|---|
| Topic jumping | Grouped sections with headers |
| Repetition | Single consolidated statement |
| Filler words/false starts | Clean prose |
| Tangents | Callouts or integrated context |
| Q&A interruptions | Blockquote dialogues |
Reorganize content by meaning, not by when things were said. A 2-hour rambling lecture about three topics becomes three clean sections, even if the speaker jumped between them.
Before writing anything:
Create a logical outline:
## [Main Topic 1]
### [Subtopic 1.1]
### [Subtopic 1.2]
---
## [Main Topic 2]
...
Use horizontal rules (---) to separate major topic shifts.
Apply these transformations systematically:
## Main Section <!-- H2 for major topics -->
### Subsection <!-- H3 for subtopics -->
#### Point or Example <!-- H4 for specific items when needed -->
Preserve speaker identities with blockquotes:
> **Participante:** ¿Cómo funciona X?
> **Instructor:** X funciona de esta manera...
For multi-turn exchanges:
> **Estudiante:** Primera pregunta
> **Profesora:** Respuesta inicial
> **Estudiante:** Pregunta de seguimiento
> **Profesora:** Respuesta expandida
| Content Type | Callout to Use |
|---|---|
| Key concept/principle | > [!important] |
| Practical advice | > [!tip] Recomendación |
| Warning/caution | > [!warning] |
| Interesting aside | > [!note] |
| Real-world example | > [!example] |
| Quoted wisdom | > [!quote] |
| Action items | > [!todo] |
| Summary | > [!abstract] or > [!tldr] |
| Success/conclusion | > [!success] |
Convert comparison discussions into tables:
| Columna 1 | Columna 2 | Columna 3 |
| --------- | --------- | --------- |
| Dato 1 | Dato 2 | Dato 3 |
When the speaker lists things (even implicitly):
- Item one
- Item two
- Sub-item
- Item three
When a process or flow is described:
```mermaid
graph LR
A[Paso 1] --> B[Paso 2]
B --> C[Paso 3]
C --> D[Resultado]
```
```python
# Example code from the presentation
def example():
return "formatted code"
```
Always include appropriate YAML frontmatter:
---