instagram-capture-process

name	instagram-capture-process
description	Process a new Instagram save (reel or carousel) into a properly-named Type B note, inject transcription, update the saved-posts index, and update the MOC. Use when [OWNER] says "instagram-capture-process", "process this reel", "process this carousel", "add this IG capture", or "process Instagram save". Outputs a verified Type B note in 06-resources/social-captures/instagram-captures/saved-posts/ with full transcription, updated index, and MOC link.
argument-hint	reel or carousel + author handle + topic slug + raw transcription (or path to source file)
allowed-tools	Read, Write, Edit, Grep, Glob, Bash
metadata	{"author":["OWNER"],"version":"1.0.0","ckis-context":true,"category":"workflow-automation"}

Instagram Capture Process

Turns a raw Instagram save (reel or carousel) into a correctly-classified, fully-transcribed Type B note — and keeps the index + MOC current. Created after G5.QC revealed that manual processing without a skill produced 61 misrouted, untranscribed notes.

━━━

Scope

This skill operates on: 06-resources/social-captures/instagram-captures/saved-posts/

Do NOT touch:

• 00-inbox/_PROFILE.md, _MEMORY.md, _INTERESTS.md, _ACTIVE-PROJECTS.md
• .obsidian/ folder
• 05-knowledge/permanent-notes/ — social captures NEVER go here (Type A only)
• Any file with status: archived in frontmatter

━━━

Pre-conditions

Before running, verify:

1. [OWNER] has provided: content type (reel or carousel), author handle, topic description, and transcription text or source file path
2. 06-resources/social-captures/instagram-captures/instagram-saved-posts-index.md exists
3. 05-knowledge/maps-of-content/MOC-Reels-Knowledge-Batch-2026-05.md exists (or ask [OWNER] which MOC to update)

If pre-conditions are not met, stop and ask [OWNER] for the missing inputs.

━━━

Phase 1: Classify and Name

1. Determine content type from [OWNER]'s input:

   • Reel (single video with audio) → prefix instagram-
   • Carousel (multi-slide static images) → prefix instagram-
   • Both use the same prefix; distinguish by subtype: social-capture and note body

2. Build the canonical filename:

   instagram-{author-slug}-{topic-slug}.md
   

   Rules:

   • author-slug: Instagram handle, lowercased, dots/underscores → hyphens
   • topic-slug: 2-5 word kebab-case topic summary
   • No special characters, no spaces
   • Example: instagram-mavenhq-agent-harnesses.md

3. Check for filename collision:

   Glob "06-resources/social-captures/instagram-captures/saved-posts/instagram-{author-slug}-*.md"
   

   If a file with the same author+topic exists, add -v2 suffix and notify [OWNER].

━━━

Phase 2: Draft the Note

4. Build the YAML frontmatter (Type B — do NOT use type: permanent-note):

---
type: resource
subtype: social-capture
processing: index-only
platform: instagram
author: {author-handle}
handle: @{author-handle}
content-type: {reel | carousel}
topic: "{topic description}"
captured: {YYYY-MM-DD}
created: {YYYY-MM-DD}
modified: {YYYY-MM-DD}
status: processed
tags: [resource, social-capture, instagram, {topic-tag}]
related:
  - "[[06-resources/social-captures/instagram-captures/instagram-saved-posts-index]]"
  - "[[05-knowledge/maps-of-content/MOC-Reels-Knowledge-Batch-2026-05]]"
---

5. Build the note body:

# {Author Handle} — {Topic Title}

> {One-sentence description of what this content covers and why it's relevant to [OWNER]/[YOUR_PROJECT].}

━━━

## Key Points

- {Bullet 1 — main insight or technique}
- {Bullet 2}
- {Bullet 3}
(3-5 bullets max — Type B means body is sacred, do not over-synthesize)

## [YOUR_PROJECT] Relevance

{1-2 sentences on how this applies to [YOUR_PROJECT] or [OWNER]'s current work.}

━━━

## Transcription

{Full verbatim transcription of the reel/carousel text content. For reels: full spoken text. For carousels: all slide text in order, with slide breaks marked as "---".}

6. Write the note to: 06-resources/social-captures/instagram-captures/saved-posts/{filename}.md

━━━

Phase 3: Update the Index

7. Read 06-resources/social-captures/instagram-captures/instagram-saved-posts-index.md to find the last row number.

8. Append one new row to the table (do not rewrite the whole file — use Edit):

| {N+1} | [[{filename-without-.md}]] | @{author-handle} | {content-type} | {captured-date} | {1-line description} |

9. Update modified: field in the index frontmatter to today's date.

━━━

Phase 4: Update the MOC

10. Read 05-knowledge/maps-of-content/MOC-Reels-Knowledge-Batch-2026-05.md (or the MOC [OWNER] specified).

11. Find the most relevant topic section for this content (e.g., "Agent Infrastructure", "Content & Personal Brand", "Sales & Positioning").

12. Append the wikilink under that section:

    - [[{filename-without-.md}]] — {1-line description}
    

13. If no existing section fits, add a new section header before the "Related MOCs" section:

    ## {New Topic Name}
    
    - [[{filename-without-.md}]] — {1-line description}
    

━━━

Phase 5: Verify

14. Run final checks:

    • File exists at correct path in saved-posts/
    • type: resource (NOT type: permanent-note)
    • processing: index-only present
    • ## Transcription section has actual content (not empty)
    • Index has the new row
    • MOC has the new wikilink
    • No .obsidian/ modifications

15. Report to [OWNER]:

    Processed: instagram-{author-slug}-{topic-slug}.md
    Type: {reel | carousel}
    Path: 06-resources/social-captures/instagram-captures/saved-posts/
    Index: row {N+1} added
    MOC: added to "{Section Name}" section
    

━━━

Examples

Example 1 — [OWNER] says "instagram-capture-process — @mavenhq posted about token savings, here's the transcript: [transcript text]":

• Phase 1: Filename → instagram-mavenhq-token-savings.md
• Phase 2: Type B frontmatter + body + full transcription
• Phase 3: Index row appended
• Phase 4: Added under "Agent Infrastructure" section in MOC
• Result: 1 verified Type B note, index +1, MOC +1

Example 2 — [OWNER] says "process this carousel — @juanaia, automatizar con Claude, slides: [slide text]":

• Phase 1: instagram-juanaia-automatizar-claude.md
• Phase 2: content-type: carousel, slide text as transcription with --- breaks
• Phase 3-4: Index + MOC updated
• Result: Complete Type B carousel note

━━━

Troubleshooting

No transcription provided: Ask [OWNER]: "What's the transcription text for this capture? I need the spoken words (reel) or slide text (carousel) before I can create the note." Do not create an empty ## Transcription section.

Author handle ambiguous: Use the exact handle from [OWNER]'s input. If unclear, ask: "What's the exact Instagram handle?" Do not guess.

MOC section unclear: If [OWNER]'s content doesn't fit any existing MOC section, surface the top 2-3 candidate sections and ask which to use.

Collision with existing file: Append -v2 to the slug and inform [OWNER]. Do not overwrite existing captures.

━━━

QA Checklist

Before marking complete:

• File in 06-resources/social-captures/instagram-captures/saved-posts/ (NOT in 05-knowledge/)
• type: resource in frontmatter
• processing: index-only in frontmatter
• ## Transcription section has real content
• Index row added with correct link and description
• MOC wikilink added in correct section
• No files deleted
• No .obsidian/ modifications