NeuraKeep
by Agentix Labsv0.1.0
Stores source-cited agent memory with review-gated updates and failure history; exposes search/read/propose tools over local stdio MCP and a CLI, plus an OAuth-gated hosted MCP endpoint.
Skills
Memory Search
Searches stored memory for relevant entries through the memory_search tool, over stdio MCP or the local HTTP server.
Memory Read
Reads a stored memory entry in full with the memory_read tool, returning its recorded content to the agent.
Explain Memory Sources
Returns the sources behind a memory entry via memory_explain_sources, so an agent can check where a stored claim came from.
Propose Memory Update
Submits a proposed change to stored memory with memory_propose_update, queuing it for review instead of writing directly.
Apply Memory Update
Commits a reviewed memory proposal with memory_apply_update, completing the review-gated write path.
Related Agents
Prismer Cloud
Gives AI agents a hosted runtime for shared error-to-fix evolution, persistent memory, compressed web context, and agen…
AgentOps for Apify Builders Bundle
Apify actor bundle for agent builders: normalize run traces for QA, control costs, and guard tool calls with a firewall…
Claw Compactor
Compresses LLM context such as code, JSON, logs, and chat history through a reversible 14-stage pipeline with no LLM ca…
Headroom
Compresses tool outputs, logs, RAG chunks, files, and history locally before they reach the model, as a Python or TypeS…