Solo: Free, then $0.01 per external tool call (1,500 free calls to start). Pro: $49/month with 10,000 included calls, then $0.01 per additional call. Scale: $199/month with 50,000 included calls, then $0.01 per additional call. Enterprise: Custom pricing with invoicing for more than 3 workspaces or
Key features
Harpe: chooses what an AI agent reads under a token budget with consistent results
Kibisis: keeps agent memory on your hardware with AES-256-GCM encryption
Argus: append-only record that detects edits with cryptographic verification
No network code by default for local and disconnected use
MIT License with published code, benchmarks and raw results
122+ managed connectors to business tools
Skills: reusable procedures for consistent AI execution
Tables: persistent records AI maintains between chats
Secure client connections without password sharing
Team profiles with per-user tool permissions
98.7% token reduction vs loading all tools at once
Works with Claude and ChatGPT through MCP protocol
What makes it different
Measured empirical validation across 6,111 questions outperforming full-context models
Zero plaintext on disk with encrypted vaults matching plaintext accuracy
78.3% fewer tokens sent to the model with a 1,200-token budget per turn
Preserves conversational chronology and relational qualifiers under token budgets
Works across 120+ tools in one MCP connection, no coding required
Encrypted credentials kept separate from AI, with granular access control
98.7% fewer tokens through on-demand tool loading and server-side queries
Perseus Computing
FloConnector