SuperLocalMemoryLocal-first Memory for AI Agents
Store context once and recall it across AI agent sessions through MCP or CLI. SuperLocalMemory keeps core memory on infrastructure you control, with workspace isolation, explicit operating modes and auditable retrieval.
Carry context across agent sessions
Project decisions and release requirements should not need to be rediscovered in every session. Store attributable context in a workspace, recall the relevant facts, and inspect their provenance. Optional external providers and integrations remain explicit choices.
Key Capabilities
Reliability Spine
Generation-fenced admission, a policy registry, verifiable memory transactions, per-projection apply/verify/compensate/erase ownership, and hash-checkable completion manifests.
Evidence-Led Retrieval
Dense, BM25 lexical, temporal, associative, and spreading-activation candidate producers fuse into a single retrieval path when their declared dependencies are healthy.
Scoped Memory and Governance
Personal, named-profile, shared, and global scopes with role-aware access, provenance, retention, export, verified erasure, and an auditable control plane.
Three Operating Modes
Mode A keeps the core local, Mode B adds a local model, and Mode C deliberately uses an external provider. Network behavior is explicit rather than hidden in the memory path.
CLI, MCP, Dashboard, and Adapters
Operate the same memory system from the command line, Model Context Protocol tools, the dashboard, and framework adapters for agent workflows.
Learning, Cache, and Compression
Feedback-driven learning, reusable retrieval/cache paths, and context compression help agents carry forward useful context without surrendering local control.
On 1 October 2026, the isolated CLI proof stored and recalled one synthetic release-checklist fact in mode A. This is a behavior check, not a retrieval benchmark. Published paper figures keep their original version, dataset and model scope.
Get Started
SkillFortify
Supply Chain Security for AI Agent Skills
Static analysis, behavioral sandboxing, and cryptographic attestation for the AI agent skill ecosystem. Detects malicious skills before they execute.
AgentAssert
Runtime Behavioral Contracts for AI Agents
Define YAML rules and check supplied structured agent state in Python. Place checks at an action boundary and raise on hard violations.
AgentAssay
Regression Testing for Non-Deterministic AI Agents
Token-efficient stochastic behavioral testing framework purpose-built for non-deterministic AI agent workflows. Part of the Qualixar AI Reliability Engineering platform.