memory_import_claude
Import Claude Code auto-memory files into AgentDB with ONNX vector embeddings. Reads ~/.claude/projects/*/memory/*.md files, parses YAML frontmatter, splits into sections, and stores with 384-dim embeddings for semantic search. Use allProjects=true to import from ALL Claude projects. Pass project...
This record as markdown: /tools/ruflo/memory-import-claude.md
What memory_import_claude does on Ruflo
AI agents use memory_import_claude to create or update resources in Ruflo, usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Ruflo environment.
Why memory_import_claude is rated Medium
The primary function is to import and store data (Claude Code auto-memory files as embeddings) into AgentDB. While it begins by reading local files, the substantive action is persisting transformed data into a database. This is reversible (data can be updated/removed later), so it's Write rather than Destructive.
From the tool's definition Tool description explicitly states it 'reads...and stores with 384-dim embeddings' into AgentDB, which is a Write operation that creates/modifies data in the agent database.
Attacks that exploit this kind of access
The rule that runs memory_import_claude safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Ruflo, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For memory_import_claude, this is the rule to start with:
memory_import_claude stays usable, but capped: an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Ruflo, apply this rule, and every memory_import_claude call is checked against it from then on.
Questions about memory_import_claude
Import Claude Code auto-memory files into AgentDB with ONNX vector embeddings. Reads ~/.claude/projects/*/memory/*.md files, parses YAML frontmatter, splits into sections, and stores with 384-dim embeddings for semantic search. Use allProjects=true to import from ALL Claude projects. Pass projectPath to override cwd-based detection (#1883 — required when Ruflo runs in WSL but Claude Code is on Windows). Pass excludeFilePatterns (glob list) or excludeFiles (absolute path list) to skip voice-load-bearing, PII, or persona-restricted files (#1937). Use when native Read/Write is wrong because you need (a) cross-session retrieval by semantic similarity (vector embeddings) not by file path, (b) namespacing across projects without managing directory layout, or (c) the .swarm/memory.db audit trail. For one-shot file I/O, native Read/Write is fine. It is categorised as a Write tool in the Ruflo MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Ruflo MCP server in PolicyLayer and add a rule for memory_import_claude: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Ruflo. Nothing to install.
memory_import_claude is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the memory_import_claude rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for memory_import_claude. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
memory_import_claude is provided by the Ruflo MCP server (ruvnet/ruflo). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Ruflo, and thousands of servers like it.
This server
Across the catalogue