SlashML (/ ml) Persistent Agent Memory & Code Localization
RemoteAbout SlashML (/ ml) Persistent Agent Memory & Code Localization
SlashML Memory is a plugin, skill and MCP server for AI coding agents, providing persistent memory, behavioral code search and bug localization. It also uses lifecycle hooks to capture standing instructions and check tool calls against active rules.
Key features
- 01Persistent cross-session memory
- 02Behavior-based repository code search
- 03Bug candidate ranking
- 04Hooks for instruction capture and rule checks
Use cases
- 01Recall project decisions in a later session
- 02Find code by its behavior
- 03Narrow down files related to a bug
- 04Apply standing rules during tool use
MCP support
The README identifies SlashML as an MCP server but does not specify its MCP connection details or exposed tools.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.slashml.ai/mcpclaude mcp add --transport http ai-slashml-ml-memory https://api.slashml.ai/mcp{"mcpServers":{"ai-slashml-ml-memory":{"type":"http","url":"https://api.slashml.ai/mcp"}}}Open in Cursor{"servers":{"ai-slashml-ml-memory":{"type":"http","url":"https://api.slashml.ai/mcp"}}}Open in VS Code{"mcpServers":{"ai-slashml-ml-memory":{"type":"http","url":"https://api.slashml.ai/mcp"}}}{"mcpServers":{"ai-slashml-ml-memory":{"serverUrl":"https://api.slashml.ai/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMESlashML Memory (slashml-memory)
Official SlashML (/ ml) plugin, skill, and MCP server for sub-millisecond persistent memory, ambient lifecycle hooks, behavioral code search, and bug localization.
Overview
SlashML provides persistent conversational memory, repository symbol recall, and automatic rule enforcement for AI coding agents:
- Persistent Memory (
search_memories,add_memory,get_memories,forget_memory): Store and retrieve architectural decisions, user preferences, and standing project rules across sessions. - Behavioral Code Search (
search_code_by_behavior,index_repository): Locate functions, classes, and modules by runtime or architectural behavior. - Bug Localization (
find_bug_candidates): Rank candidate functions and files most likely responsible for a symptom or stack trace. - Ambient Lifecycle Hooks (
hooks/hooks.json): Automatically assimilate new standing instructions onUserPromptSubmitand check tool calls against active rules onPreToolUse,PostToolUse, andStop.
Configuration
Install directly in Claude Code:
/plugin install slashml-memory --marketplace slashml-ai/ml-memory
When installing the plugin in Claude Code, provide your SlashML API key (sml_live_... from https://app.slashml.ai) when prompted for the api_key plugin option.
Python & TypeScript SDK Usage
Python (slashml-memory)
from slashml import SlashMLMemory
memory = SlashMLMemory(api_key="sml_live_...")
memory.add("User prefers dark mode and strict TypeScript.", user_id="u_123")
hits = memory.search("What are the user's coding preferences?", user_id="u_123")
TypeScript (@slashml/memory)
import { SlashMLMemory } from "@slashml/memory";
const memory = new SlashMLMemory({ apiKey: "sml_live_..." });
await memory.add("We migrated auth from JWT to session cookies", { userId: "team_backend" });
const results = await memory.search("How does auth work?", { userId: "team_backend" });
License
Apache-2.0. See LICENSE for details.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.