Delete a specific past task from the project's memory to erase sensitive or incorrect lessons, permanently removing its executions, tool calls, and experience nodes from the local Experience Graph.
MIT
Destructive
597,353 tools. Updated 2026-09-21 14:35
"Designing a memory system for persona-based agents to improve user experience" matching MCP tools:
- Retrieve the current persona and user context at session start to load the AI's identity and user knowledge for memory continuity across sessions.MIT
- Retrieve a complete system prompt for a persona, including constitution and UVC qualities, to enforce AI safety guardrails.MIT
- Discover all users and agents with stored memories and their memory counts to identify system participants before conducting searches.MIT
- Retrieve host system details including operating system, CPU, memory, user, and active workspace for environment inspection.MIT
- Export LoreConvo sessions to Anthropic managed-agents memory format for import into memory stores. Filters sessions by project, IDs, or days back.Business Source 1.1
Matching MCP Servers
- AlicenseAqualityBmaintenanceMCP server for token-user-system wallet operations, supporting balance, trend, transactions, coupons, and corporate transfer info.542 npmMIT
- FlicenseNot gradedqualityBmaintenanceStores photos and notes as experience memories in Google Drive and enables natural language search to retrieve them later.-
Matching MCP Connectors
Orbit is an MCP server that gives your AI agents a shared memory layer and live dashboard, every agent reads active decisions, logs outputs to a searchable vault, and reports status in real time. 2-minute setup via MCP; works across Claude, ChatGPT, Codex, Cursor, Gemini, and Manus.
Decision Layer for AI Agents — 58+ tools, Advisor, MCP. Free key: POST /v1/register {}.
- Retrieve a configured driver persona by name to access its full configuration, including the persona prompt, for use in evaluations and simulations.Apache 2.0
- Retrieves detailed information about an AI persona using its unique ID. Provides persona details for integration with Anam AI.MIT
- Fetches the agent's assembled persona and system prompt from Superpos, including SOUL, RULES, STYLE, MEMORY, pre-assembled server-side.MIT
- Retrieve the desired, disliked, and never qualities for a specified persona from the UVC system, using the provided persona ID.MIT
- Share a memory with specific agents to provide them with relevant context for coordinated and accountable actions.MIT
- Generate a share link for a persona with optional expiration and usage limits to control access.MIT
- Evaluate user requests against active constitutions and persona to generate a policy decision kernel.MIT
- Register a Claude persona on AgentDNS and start a background A2A server, enabling other agents to discover and communicate with this agent.MIT
- Store explicit user preferences and query-rewrite corrections in persistent memory so future interactions honor clarified intent and improve retrieval accuracy across sessions.MIT
- Create a configurable AI agent persona for Ethora chat, defining its name, avatar, system prompt, and reply behavior (response mode, cooldown) for use in single or multi-agent rooms.MIT
- Share a memory to make it accessible to all agents serving the same user or the entire organization. Use to propagate valuable knowledge like user preferences or team conventions across agents.Apache 2.0
- Confirm a pending memory after user approval to activate it in long-term memory. For corrections, preferences, or SOPs, also refresh the user persona.MIT
- Saves or overwrites persona markdown to define behavior rules, supporting project-specific or global scope selection for the base persona.MIT
- Confirm a pending behavior correction to permanently apply it, writing the correction to memory and refreshing the persona.MIT
- Retrieve per-host memory usage in GB for a given time window. Default for ambiguous server or system memory questions.MIT