AI Plugin Indextoken-savior vs Caveman Shrink
token-savior and Caveman Shrink are both knowledge and memory MCP servers. Caveman Shrink scores higher (94 vs 88 out of 100). Caveman Shrink has more GitHub stars (114k), and Caveman Shrink was updated more recently (yesterday). The table compares them side by side.
| token-savior | Caveman Shrink | |
|---|---|---|
| Score | 88 / 100 | 94 / 100 |
| What it does | Structural codebase MCP server with persistent memory: navigate by symbol, recall across sessions. | MCP proxy that compresses prose fields (tool descriptions, etc.) using caveman rules. Same accuracy, fewer context tokens. |
| Publisher | Community: Mibayy | Verified: npm:caveman-shrink |
| Grades | Maintenance A, Licence A, Quality A | Maintenance A, Licence A, Quality A |
| npm downloads a week | Not on npm | 1,525 |
| Runs | Local | Local |
| Works with | Claude Code, Claude Desktop, Cursor, Codex, Gemini CLI, VS Code, Windsurf | Claude Code, Claude Desktop, Cursor, Codex, Gemini CLI, VS Code, Windsurf |
| GitHub stars | 1,182 | 113,764 |
| Last commit | 15 Oct 2026 | 19 Oct 2026 |
| Licence | MIT | Apache-2.0 |
| Listed since | 19 Oct 2026 | 20 Oct 2026 |
| Install token-savior | Install Caveman Shrink |
Questions
Is token-savior or Caveman Shrink more popular?
Caveman Shrink has more GitHub stars: 113,764.
Can I use token-savior and Caveman Shrink together?
Yes. AI assistants can load several MCP servers at once. Install both and the assistant picks the right one for each task.