A
licenseNot graded
qualityD
maintenanceA portable MCP server providing a shared intelligent memory system for any MCP-compatible AI tool, enabling storage, retrieval, extraction, and governance of memories across sessions.
18 npm
MIT
mymemory-mcp
by 20vision
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mymemory-mcpfetch my memory context"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mymemory-mcp
One memory for every AI. MyMemory is a user-owned, portable memory vault: import your context once (CLAUDE.md, custom instructions, bios), and every AI tool you use reads the same distilled entries. Agents propose new memories after conversations; proposals land in a pending queue you promote or reject. The memory belongs to you, not a lab. This package is the vault as an MCP server: 3 tools, zero install beyond npm.
Your context, portable. — myclawn.com/mymemory
Quick start
Get an API key first: open https://www.myclawn.com/mymemory, sign in with a
passkey, and create a key in the Keys view. Keys look like mm_… and
are shown once — the read + propose scopes cover everything below.
Claude Code
claude mcp add mymemory -e MYMEMORY_API_KEY=mm_your_key -- npx -y mymemory-mcpClaude Desktop (claude_desktop_config.json)
{
"mcpServers": {
"mymemory": {
"command": "npx",
"args": ["-y", "mymemory-mcp"],
"env": {
"MYMEMORY_API_KEY": "mm_your_key"
}
}
}
}Any other MCP client: serve over stdio with command npx and args
["-y", "mymemory-mcp"]; the key goes in the MYMEMORY_API_KEY env var.
Self-hosted or testing against a local app? Point MYMEMORY_BASE_URL at
the origin (default https://www.myclawn.com).
Related MCP server: SOMA MCP
What your agent can do
Tool | What |
| Fetch the vault's compiled context block (directives first, then facts/preferences/notes) plus the raw entries — call once at conversation start |
| Case-insensitive substring search over active entries ( |
| Propose up to 20 |
The API key's scopes gate the verbs: read covers context + search,
propose covers proposing. A read-only key proposing gets a clear 403.
What it deliberately does NOT do
No approving/rejecting entries, no editing the vault, no minting or revoking keys — curation stays with the human at myclawn.com/mymemory, where every proposed entry waits in a review queue. This package is the read/propose surface: agents remember, you decide.
Links
The vault app: https://www.myclawn.com/mymemory
API + pairing surface:
https://www.myclawn.com/api/mymemory/*iOS app source (pairs via 6-char code):
ios/in the monorepo
Decisions
Search is client-side. The API contract defines no search route, so
mymemory_searchfetches/api/mymemory/contextonce and filters entries by case-insensitive substring overtext(andkind, so "directive" lists all rules). Simple and fine at vault sizes.mymemory_get_contextreturns text-first. The compiledcontext_blockleads the tool result (paste-ready), followed by the raw entries as JSON for structured use.MYMEMORY_BASE_URLis the origin, not the full API prefix; routes are appended as/api/mymemory/…. Trailing slashes are tolerated.Engines
node >=22per the build contract (siblingmyclawn-mcpsays >=20; the contract wins).Registry metadata (
server.json/smithery.yaml) ships in this package; submission steps live indocs/registry-submissions.md. The package is publish-ready (publishConfig.access: public).
MIT © MyClawn
Available Tools
3 toolsmymemory_get_contextA
Fetch the user's MyMemory vault context: a compiled context_block (directives first, then facts/preferences/notes, one per line) plus the raw active entries. Call once at conversation start and apply it — these are the user's standing rules, facts, and preferences across every AI tool they use.
ParametersJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
TDQS
A4.6/5.0
Behavior4/5
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It discloses the content structure (context_block format, directives-first ordering) which is genuinely useful behavioral info. It doesn't mention authentication requirements, pagination, or failure modes, but for a zero-parameter read-only fetch tool the disclosure is reasonably complete for what's exposed.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Conciseness4/5
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, front-loaded with purpose then structure then usage guidance. Slightly dense single paragraph but every clause earns its place. Could optionally be split into bullet points but it's efficient as written.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Completeness5/5
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
A zero-parameter, no-output-schema tool with clear READ semantics. The description covers what it returns, the internal structure, when to call it, and what to do with the result. For a simple fetch tool this is fully complete — nothing meaningful is left unexplained.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Parameters4/5
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
There are 0 parameters and schema coverage is 100%, making parameter semantics a non-issue. The dimension baseline for 0 params is 4, and the description adds context about what behavior to expect without parameters (returns full context).
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Purpose5/5
Does the description clearly state what the tool does and how it differs from similar tools?
Clear verb (Fetch) + specific resource (user's MyMemory vault context) with detailed description of what's returned (compiled context_block with directives/facts/preferences/notes structure plus raw active entries). Distinguishes well from siblings: get_context returns the full compiled context while search/propose serve different purposes.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Usage Guidelines5/5
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly states 'Call once at conversation start and apply it' — gives a precise trigger condition. The description's ending ('these are the user's standing rules, facts, and preferences across every AI tool they use') explains why this tool matters and implies it should be prioritized early, which differentiates it from alternative sibling tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
mymemory_proposeA
Propose new entries to the user's MyMemory vault when you learn something durable about them. Entries land in a PENDING queue the human reviews in the app — nothing becomes active without their approval. Only propose lasting cross-session knowledge (a rule, fact, preference, or note about the user), never one-off task details or things already in the vault. kinds: directive = a rule ("never …", "always …"), fact, preference, note. Max 20 per call; the server dedupes against existing entries.
ParametersJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| entries | Yes |
TDQS
A4.3/5.0
Behavior4/5
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. It effectively discloses that entries go to a PENDING queue, that nothing activates without human approval, that there's a max of 20 per call, and that the server dedupes. This is strong behavioral transparency for a write tool with no annotation support.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Conciseness5/5
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single cohesive paragraph that front-loads the core purpose, then layers in constraints in logical order — durable-only scope, kind taxonomy, limits, dedupe behavior. Every sentence earns its place, with no wasted words or redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Completeness4/5
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a single-parameter write tool with no output schema, the description covers the essential usage context well: what to propose, what not to propose, the kinds, limits, and review flow. The only minor gap is that it doesn't specify return confirmation behavior, but this is less critical given no output schema is expected.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Parameters4/5
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate for the undocumented 'entries' parameter. It does by explaining what kinds are available (directive, fact, preference, note with examples for directive) and explaining the dedupe and pending-queue behavior. The description adds meaning beyond the bare schema, though it doesn't detail the text length constraints or per-kind examples exhaustively.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Purpose4/5
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states what the tool does: propose new entries to the user's MyMemory vault with a specific verb ('propose') and resource ('MyMemory vault'). It distinguishes from siblings by clarifying this is about writing new durable knowledge, while get_context and search are likely read operations. However, it doesn't explicitly contrast against the siblings, so it's clear but not fully differentiated within the sibling set.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Usage Guidelines5/5
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit when-to-use guidance ('when you learn something durable about them'), what NOT to propose ('never one-off task details or things already in the vault'), and the kinds taxonomy. It also explains the pending-queue approval model. This is exemplary usage guidance, clearly stating scope and exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
mymemory_searchA
Search the user's MyMemory vault for a word or phrase. Case-insensitive substring match over active entries (text and kind). Returns { query, count, matches }.
ParametersJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Word or phrase to look for, e.g. "diet", "timezone", "typescript" |
TDQS
A4.2/5.0
Behavior4/5
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It discloses that matching is case-insensitive, substring-based, limited to active entries (not inactive/deleted), and searches both text and kind fields. It also states the return shape ({ query, count, matches }). This is solid behavioral disclosure for a read/search tool, though it doesn't mention pagination or match limits which could matter for large vaults.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Conciseness5/5
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, zero wasted words. Every clause adds value: scope (user's vault), function (search for word/phrase), match semantics (case-insensitive substring), search surface (text and kind), filter (active entries), and return shape. Perfectly front-loaded with the most important info first.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Completeness4/5
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple 1-parameter read tool with 100% schema coverage and clear return description, this is nearly complete. The only gap is no mention of result limits or pagination behavior for large match sets, and no explicit output schema. But given the tool's simplicity, the description is adequate to allow correct invocation. Minor deduction for the missing result-limit detail.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Parameters3/5
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% - the single query parameter is fully documented in the schema with description and examples. The description adds the case-insensitive substring detail and the scope (active entries, text and kind fields), which enhances the schema. This lands the baseline at 3 since the schema already does the heavy lifting, and the description adds modest enrichment.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Purpose5/5
Does the description clearly state what the tool does and how it differs from similar tools?
Description uses specific verb+resource: 'Search the user's MyMemory vault for a word or phrase.' It clearly states the matching behavior (case-insensitive substring) and scope ('active entries (text and kind)'). This distinguishes it from siblings mymemory_get_context and mymemory_propose, which presumably retrieve context or propose additions rather than search.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Usage Guidelines4/5
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains the matching semantics (case-insensitive substring over active entries, both text and kind fields) which gives clear context on what the search will return. It doesn't explicitly state when NOT to use it or name alternatives, but the scope restriction to 'active entries' and the clear verb distinguishes usage from siblings. Slight deduction for no explicit exclusion guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
3 tool updates
v0.1.0- First observed
mymemory_get_context - First observed
mymemory_propose - First observed
mymemory_search
TDQS
A4.1/5.0
Scored across 3 tools
Disambiguation4/5
The three tools have distinct purposes: get_context (retrieve full vault), search (find specific entries), and propose (add new entries). While get_context and search both read data, their purposes are clearly separated (bulk fetch vs. targeted lookup), and propos e is clearly a write operation. No meaningful ambiguity between them.
Naming Consistency5/5
All tool names follow a consistent mymemory_<verb> pattern: get_context, search, propose. The verbs are uniform and descriptive, and each name clearly signals its operation (retrieval, lookup, and write). The mymemory_ prefix consistently scopes the namespace.
Tool Count3/5
Three tools is on the thin side but understandable for a personal memory vault: fetch, search, and write. The scope is narrow enough that three tools could work, though one might expect an explicit update/delete mechanism. It's at the lower boundary of reasonable.
Completeness3/5
The vault supports retrieval (get_context), search, and creation (propose, gated by PENDING approval). However, there's no tool to update or delete existing active entries, nor to approve the pending queue, meaning maintenance of the vault is delegated entirely to the human in the app. This leaves some lifecycle gaps beyond the core read/write flow.
Maintenance
ActivitySlowing
ResponsivenessNo issues
Related MCP Connectors
Person-owned AI memory that learns, not just stores — portable context for any MCP client.
Persistent, portable memory for AI assistants — your private memory graph, from any MCP client.
- JustOnceOAuthai.justonce
Persistent memory for AI assistants — one shared, OAuth-secured vault for every MCP client.
An MCP memory server. One memory your agents share — across models, devices and apps.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceA private, self-hosted MCP server that wraps a retrieval pipeline over your own data, enabling trusted AI agents to access and manage your personal memory through standard MCP tools.Apache 2.0
- AlicenseNot gradedqualityDmaintenanceA portable self-hosted memory layer for AI tools, storing context, memories, and handoffs for access from any MCP-compatible client.17 npmMIT
- FlicenseNot gradedqualityCmaintenanceA personal memory MCP server that stores and retrieves conversation memories, enabling AI agents to recall past discussions, promises, and preferences using natural language.-