Agent Chatroom Ledger

See the idea in action.

After restart: Analyst has 0 unread, Reviewer has 2; excluding message 1 leaves canonical IDs [1, 2] and visible ID [2].

Two synthetic messages, read independently by Analyst and Reviewer.

{
  "unread_before": {
    "Analyst": 2,
    "Reviewer": 2
  },
  "unread_after_restart": {
    "Analyst": 0,
    "Reviewer": 2
  },
  "canonical_ids": [
    1,
    2
  ],
  "visible_ids_after_exclusion": [
    2
  ],
  "canonical_lens_survived_broken_view": true,
  "rebuilt_visible_ids": [
    2
  ],
  "reviewer_unread_after_explicit_acknowledgement": 0,
  "verification": {
    "ok": true,
    "error_count": 0,
    "issues": []
  }
}

Try the example.

From the repository root, follow the dependency requirements in the README. This example uses supplied synthetic material.

python -m pip install -e .
python -m examples.walkthrough

After restart: Analyst has 0 unread, Reviewer has 2; excluding message 1 leaves canonical IDs [1, 2] and visible ID [2].

The interesting part.

Canonical messages survive changes to compiled context views and per-consumer cursors.

• Runnable example and its input
• Complete captured result
• agent_chatroom_ledger/service.py
• agent_chatroom_ledger/store.py
• agent_chatroom_ledger/cli.py
• tests/test_chatroom_backend.py