| name | investor-memory-sync |
| version | 1.0.0 |
| description | Batch-sync enriched investor profiles from intel/ into Claude memory for
cross-session recall. Creates condensed memory files and updates MEMORY.md index.
Run after enriching investors to make intel available in all future conversations.
|
| allowed-tools | ["Bash","Read","Grep","Glob","Write"] |
Investor Memory Sync — Persist Intel to Claude Memory
When to Use
- "add investors to memory", "sync investors", "push intel to memory"
- After running investor enrichment batch
Workflow
Step 1 — Scan Intel Directory
Read all profiles from:
intel/investors/*.mdintel/contacts/*.mdintel/developers/*.md
Step 2 — Create Memory Files
For each profile, create a CONDENSED memory file at the global memory directory:
~/.claude/projects/-Users-ilessio/memory/
This is the global memory — loaded in ALL sessions (Claude Code + Claude Desktop).
Do NOT use project-specific paths like -Users-ilessio-Development-*.
Format:
---
name: investor-{slug}
description: {Name} — {Role} @ {Fund} — {type} profile for BD intelligence
type: reference
---
**Key Info:**
- Telegram: @handle
- LinkedIn: URL
- Fund: Name
- Role: Title
- Stage: X
- Sectors: X, Y, Z
- Ticket: $Xk-$XM
**BD Status:** not contacted | contacted | in conversation | closed
**Approach:** One line on how to approach them
**Last Updated:** YYYY-MM-DD
Step 3 — Update MEMORY.md Index
Add entries grouped under ## Investor Intelligence. Keep under 200 lines.
If full, create sub-index: memory/INVESTORS-INDEX.md
Step 4 — Batch Processing
Process in batches of 10. Report progress after each batch.
Step 5 — Summary
Report totals, skips, and remind to run /push-claude.
Rules
- Memory files are CONDENSED — not full copies
- Intel files in
intel/ are the source of truth (detailed)
- Memory files in
memory/ are for quick Claude context
- Check for existing files before creating duplicates
- Skip group admins
