Search past conversation memories by meaning to recall decisions and context across sessions. Enter a natural-language query to retrieve relevant entries when recency is not enough.
MIT
524,660 tools. Updated 2026-09-06 17:35
"Finding the Best Tool for Memory Context Across Agentic Sessions Using Augment Code" matching MCP tools:
- Store failed approaches, architecture decisions, and setup gotchas as memory for future retrieval. Use when git would not surface key context needed in later sessions.MIT
- Initialize a versioned, venture-scoped context filesystem to persist plans, decisions, and code snippets across sessions. Safe to call repeatedly; required once before storing context.MIT
- Replace the existing per-friend memory blob to store persistent facts across sessions. Retrieve current memory first to merge updates.Apache 2.0
- Restore session context by loading prior session memory—hot topics, open items, protocols, and memory rules—so follow-ups stay accurate; call at the start of every conversation to avoid losing context.MIT
- Save session summaries, decisions, and artifacts to persistent memory for recall in future sessions. Ensures context continuity across AI agent interactions.Business Source 1.1
Matching MCP Servers
- AlicenseAqualityBmaintenanceMCP server for persistent, cross-session, local-first memory for AI agents, storing memories as Markdown files with SQLite indexing for hybrid search.24Apache 2.0
- AlicenseNot gradedqualityAmaintenanceEnables AI coding agents to collaborate on the same project by sharing session briefs and reading each other's native transcripts, memories, and instructions in place, with zero-copy, across different agent tools.58MIT
Matching MCP Connectors
LinkdAPI - the best linkedin professional data API: Powerful, lightweight B2B Data API for.
Cloudflare Workers MCP server: code-explainer
- Store a key-value pair in the agent's encrypted, persistent memory. Survives across sessions.MIT
- Save key-value pairs to persistent memory across sessions for remembering user preferences, installed skills, or project context.MIT
- Persist project conventions, decisions, and gotchas to encrypted memory, ensuring knowledge survives across sessions. Use stable keys to update existing memories.MIT
- Save facts, documents, code, or user details into persistent memory for cross-session retrieval. Use for additive, non-destructive storage without context loss.Apache 2.0
- Fetch relevance-ranked memories using natural-language queries. Filter by memory type or path and optionally prioritize working memory for quick, best-effort context.MIT
- Save facts, observations, or context to persistent memory for recall across sessions. Optionally categorize entries for organized retrieval.MIT
- Add knowledge entries like code examples and best practices to the Klever VM context store for later retrieval via query_context or search_documentation.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
- Set the default memory repository for the plugin, persisting the selection across sessions within the same installation.MIT
- Save durable facts, preferences, or decisions to persistent memory that works across sessions and AI tools. Ideal for content that must survive across chats and projects.MIT
- Store durable user facts and preferences after responding to maintain context across sessions without SDK integration.MIT
- Log a new memory with specific details like text, category, client ID, and importance to build a searchable personal knowledge base across sessions.MIT
- Store text in a named context slot to preserve key information across agent sessions. Use it to save project state, user preferences, or plans that must survive restarts.MIT
- Manage compact project memory across chats, agents, and IDEs: record active work, complete and archive sessions, search history, and update durable rules.MIT
- Select the best tool or skill for a given task using server-side routing config. Returns the chosen option and candidates without executing anything.MIT