"author:GhostPawJS" matching MCP servers:

Enables agents to capture, shape, schedule, search, and retire tasks, deadlines, calendar events, and recurring commitments through twelve intent-shaped tools backed by a single unified SQLite model. Every call returns structured, discriminated results with next-step hints instead of thrown exceptions or ambiguous prose.

Enables agents to create and evolve evidence-backed cognitive identities by storing narrative essences, traits, and attributed observations, then refining, leveling up, and rolling back those identities. It also renders stable identity blocks for use as system prompts and exposes tools for searching souls, managing traits, and discovering templates.

Standalone belief and memory engine for Node.js on SQLite. Zero dependencies, built-in LLM tool facade.

Enables LLM agents to query and full-text search a local SQLite mirror of Telegram bot messages, chats, users, reactions, reply chains and stored file blobs, plus send, edit, pin and broadcast messages or ban, restrict and promote members. Without a bot token it returns action plans, while local reads and tool discovery work offline.

Enables LLM agents to manage a local SQLite-backed social CRM through twelve intention-shaped tools: searching and inspecting contacts, merging duplicates, managing identities and relationships, recording events and commitments, and maintaining recurring dates, attributes, and derived graph review views.