playbooks

home / skills / openclaw / skills / transcript-triage

transcript-triage skill

/skills/nonlinear/transcript-triage

This skill triages long transcripts by extracting topics, decisions, and actions, then organizes them into prioritized roadmap-worthy items for your notes.

npx playbooks add skill openclaw/skills --skill transcript-triage

Review the files below or copy the command above to add this skill to your agents.

Files (2)
SKILL.md
1.9 KB
---
name: transcript-triage
description: "Parses long transcripts into topics under ordered lists, to triage in your notes system"
type: public
version: 0.1.0
status: published
dependencies: []
author: nonlinear
license: MIT
---

## Transcript Triage

**Parse AI conversation transcripts → numbered topics → groom into positions.**

### Flow

1. **Trigger:** User provides transcript (MD format)
2. **Parse:** AI extracts topics, decisions, action items
3. **Triage:** Classify by urgency/importance:
   - **NOW** - Current epic (immediate)
   - **LATER** - Future epics (planned)
   - **BACKLOG** - Ideas (someday/maybe)
   - **DECISIONS** - What was agreed (log)
4. **Groom:** User positions topics in correct places (ROADMAP, epic-notes, memory)

### Triggers

- "triage transcript"
- "parse conversation"
- "organize chat"
- "extract action items from [file/paste]"

### Output Format

```markdown
## Triage Results

### NOW (v1.1.0 Contract Diagram)
1. [Topic] - [Brief description]
2. [Topic] - [Brief description]

### LATER (Future Epics)
3. [Topic] - [Brief description]
4. [Topic] - [Brief description]

### BACKLOG (Ideas)
5. [Topic] - [Brief description]

### DECISIONS (Log to Memory)
- [Decision] - [Context]
- [Decision] - [Context]
```

### Integration Points

- **Backstage:** Auto-add NOW items to current epic-notes/
- **ROADMAP:** Suggest LATER items as new epics
- **Memory:** Log DECISIONS to memory/YYYY-MM-DD.md

---

## Notes

**Why this skill?**
- External AI chats (Claude, ChatGPT, etc.) generate valuable insights
- Transcripts are hard to parse manually
- Need systematic way to extract → classify → act

**Similar skills:**
- Backstage (epic management, checks)
- Roadmap (epic planning, grooming)
- Memory (decision logging)

**Unique value:**
- Bridges external conversations → internal workflow
- Automates triage (urgency + importance)
- Preserves context (decisions logged, not lost)

Overview

This skill parses long AI or meeting transcripts and converts them into ordered, actionable topics for triage in your notes system. It extracts topics, decisions, and action items, then classifies them by urgency so you can quickly route items into roadmap, epic notes, or memory. The result is a clean, numbered list ready for grooming and execution.

How this skill works

You provide a transcript (Markdown or pasted text). The skill scans the conversation, identifies topics, decisions, and action items, and groups them into four categories: NOW, LATER, BACKLOG, and DECISIONS. It outputs a numbered, Markdown-friendly triage that you can drop into your notes, roadmap, or decision log and offers simple mappings for integration with epic-notes and memory files.

When to use it

  • After long AI chat transcripts (ChatGPT, Claude, etc.)
  • Following recorded meetings or interview transcripts
  • When importing external conversations into your planning system
  • Before grooming sessions to pre-sort potential epics
  • When you need to preserve decisions and context for future reference

Best practices

  • Provide transcripts in Markdown or clearly delimited text for best parsing
  • Include speaker labels and brief timestamps to preserve context
  • Review NOW items immediately and assign owners before moving to roadmap
  • Use LATER items to propose new epics during planning sessions
  • Log DECISIONS to dated memory files to maintain a searchable history

Example use cases

  • Convert a client kickoff transcript into immediate action items and backlog ideas
  • Triage a cross-team chat to populate the current sprint epic-notes with NOW tasks
  • Extract product decisions from an external consultant chat and add them to the memory log
  • Turn a research interview transcript into prioritized roadmap opportunities
  • Run batch triage of archived conversations to surface buried epics

FAQ

What input formats are supported?

Primarily Markdown transcripts or plain text with clear speaker labels and punctuation for accurate parsing.

How are urgency categories defined?

NOW is immediate/current epic work, LATER are planned future epics, BACKLOG holds ideas/someday items, and DECISIONS captures agreed outcomes to log into memory.