Idea creating tier1 documents - robbiemu/aclarai GitHub Wiki

โœ… Create Tier 1 Markdown files with aclarai:id comments and Obsidian ^anchors

What it does:

Converts a conversation (list of utterances) into a Markdown file where each utterance is:

  • A speaker: text block

  • Followed by:

    <!-- aclarai:id=blk_xyz ver=1 -->
    ^blk_xyz
    

This allows:

  • Precise linking between Markdown and graph nodes
  • File-safe sync and version tracking
  • Native compatibility with Obsidian block references

๐Ÿง  Context

We are generating a Tier 1 document โ€” this is the raw transcript layer. It represents the conversation as spoken, with minimal transformation.


๐Ÿ’ฌ Example Conversation Input

[
  { "speaker": "Alice", "text": "Let's release v1.2 next week." },
  { "speaker": "Bob", "text": "Sounds good. Should we announce it on Friday?" },
  { "speaker": "Alice", "text": "Yes, I'll prep the release notes." }
]

๐Ÿ“ Tier 1 Markdown Output

Alice: Let's release v1.2 next week.
<!-- aclarai:id=blk_1a2b3c ver=1 -->
^blk_1a2b3c

Bob: Sounds good. Should we announce it on Friday?
<!-- aclarai:id=blk_4d5e6f ver=1 -->
^blk_4d5e6f

Alice: Yes, I'll prep the release notes.
<!-- aclarai:id=blk_7g8h9i ver=1 -->
^blk_7g8h9i

๐Ÿ“Œ Details

  • Each utterance is one block, identified by a unique aclarai:id
  • ver=1 indicates it's the first version
  • Obsidianโ€™s ^anchor is aligned to the block ID
  • These anchors enable linking from Tier 2 summaries directly back to Tier 1
  • This file is written into the Tier 1 vault directory as something like: tier1/2024-05-20_team_chat.md

Implementation note:

IDs are UUIDs or hashes; ver=1 starts at first write.