VaultCRM - Todos

Things to solve

• Onboarding
  • context
    • personal ⇒ Not synced (gitignore)
      • user.md
      • soul.md
      • communication-style.md ⇒ contains the communication style.
      • (any other context user might want to add that’s not synced)
    • shared (shared context which is shared(added and commited to git))
      • icp.md
      • organisation.md
      • playbooks ⇒ (new folder has all Unstructured files, research, playbooks, or general context)
        • methodology.md
        • brand-communications-guidelines.md ⇒ (optional) useful for communication style umbrella
        • objection-handling.md (optional) ⇒ useful for prepping for calls and responding/followup up
        • `qualification-handling.md
        • other files
  • notes (folder)
    • daily → (folder) place for daily scheduled notes
    • weekly → (folder - optional) place for weekly scheduled notes
    • monthly → (folder - optional) place for monthly notes.
  • Starting point
    • Upload documents / import Linkedin
    • Start from an existing repo
• To solve
  • Emails and Calendar
  • Cron or equivalent.
• Templates
  • References
    • https://github.com/thephm/hal_md?tab=readme-ov-file
  • person
    • Name / email / phone / reddit
    • last contacted
    • Quick profile
    • Career History
    • Linkedin Activity
    • Contact and connection
    • Relationship notes
    • Activity Logs
• Context engineering
  • https://grok.com/c/fde4d82e-385c-450f-9ea4-afe9b327e1ea?rid=5340217c-1918-4f98-baaa-9ccbe72dde56
  • agents.md
  • heartbeat.md equivalent?
  • update_records.md ⇒ makes sure that every fetch or action is updating ALL the relevant records in the CRM
    • person / company / deal ⇒ CURD
      • Activity history
      • update last contacted
    • daily notes
      • Todos
    • memory.md
    • Agent activity history
• Teams
  • Which files to sync, which ones to not.
  • Sync as a feature later
• Technical Debt
  • Github back up should be vault specific
  • Make opening codex/open claw/claude configurable.
  • Symlinks or ID based indexing ⇒ Prevents the name change and other things to break the link
  • indexes → if the list becomes way too big what happens
    • JSON index
    • SQL Lite
    • vector db
  • Concurrency → shared updates (teams and Agent)
  • Schema Drift
    • pydantic validation