Query Markup Documents - On-device hybrid search for markdown files with BM25, vector search, and LLM reranking
Memoria is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
What it does
Local-first memory for AI coding tools: recall decisions, tasks and fixes across sessions.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user xinzzu-memoria-mcp '--' npx -y @xinzzu/memoria mcp
# 2 — ask your agent for something these tools do
# recall, project_context, file_context and 6 more
Memoria costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Where are you running it?
Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Memoria as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.
This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.
claude mcp add --scope user xinzzu-memoria-mcp '--' npx -y @xinzzu/memoria mcpReconnect, or start a new session, and the tools appear in the model’s tool list.
One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.
9 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
Read from its source code. A server that builds its tool list when it starts can list more tools than this.
Search memories of previous work by keywords/question. Returns facts, decisions, tasks and fixed errors. Call at session start!
| Parameter | Type | Required | Notes |
|---|---|---|---|
| query | string | yes | Question or keywords, e.g. "why did the build fail" |
| limit | number | no | Maximum number of results |
| project | string | no | Optional — limit to one project (path or name) |
| maxChars | number | no | Optional — cut the result at N characters to fit the context window |
| mode | string | no | "index" = one line per hit with id; fetch details with get_memory |
Fetch the latest context of a project: decisions, pending tasks, completed work. Default mode "index" returns one compact line per memory with its id (cheap on tokens); call get_memory with the ids you need for full text. Use mode "full" to get everything at once.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| project | string | yes | Project path or name |
| limit | number | no | Maximum number of memories |
| mode | string | no | "index" = one line per memory with id; "full" = complete text |
Before editing a file, ask what is known about it: fixed errors, decisions, open tasks and preferences that mention this file path. Pass an absolute or relative path.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | File path, e.g. src/engine/http.ts |
| project | string | no | Optional — limit to one project (path or name) |
| limit | number | no | Maximum number of results |
Fetch the full text of specific memories by id (ids come from project_context, recall, or related).
| Parameter | Type | Required | Notes |
|---|---|---|---|
| ids | array | yes | Memory ids (max 20) |
Store an important memory (decision, fact, user preference, completed/pending task) so every tool remembers it.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| kind | string | yes | Memory type: decision, fact, preference, open task (task_pending), finished task (task_done), fixed error (error_fixed) or general context |
| content | string | yes | Memory content, short and dense |
| project | string | no | Optional — project path or name; omit for a global memory |
| importance | number | no | 0 to 1; higher ranks first in context |
Fetch memories connected (backlinks/similarity) to one memory. Use when recall context is incomplete or you need related detail. Pass a memory id or a text query.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| id | number | no | Memory id from recall or project_context |
| query | string | no | Or keywords: the closest memory is found first, then its connections are returned |
Correct an existing memory: replace its content (auto re-embedded so semantic search stays accurate), change its kind, or adjust importance. Pass the memory id from recall.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| id | number | yes | Memory id from recall |
| content | string | no | Replacement content, short and dense |
| kind | string | no | Memory type: decision, fact, preference, open task (task_pending), finished task (task_done), fixed error (error_fixed) or general context |
| importance | number | no | New importance, 0 to 1 |
Delete a stale or wrong memory. Pass the memory id for a direct delete, or a query: a confident match is deleted immediately, otherwise candidates are listed for you to pick an explicit id.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| id | number | no | Memory id (direct delete) |
| query | string | no | Keywords to find the memory to forget |
One ready-to-use context block for a project: auto-maintained project brief + key memories + optional query matches + 1-hop relations, capped to maxChars. Ideal first call at session start.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| project | string | yes | Project path or name |
| query | string | no | Optional focus — matching memories are added to the block |
| maxChars | number | no | Upper bound on the block size in characters |
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under MIT
- Runs on your machine — nothing of it reaches our gateway
- Nothing to cap, because mcprush bills nothing
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install Memoria or to call it, at any volume.
Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| verification-discount | source | publisher verification (provenance) — cryptographic build provenance ties the artifact to its source | −0 |
| coverage-honesty | source | inspection depth (source) — how much of the target the scan could see | −0 |
What the scan could actually read
A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.
Tools were statically extracted from the published source (9 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
Capability — what it could do if the model were manipulated
Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. high is the level these add up to.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| recall | no tags | |
| project_context | untrusted-input | untrusted-input: keyword "fetch" |
| file_context | no tags | |
| get_memory | untrusted-input | untrusted-input: keyword "fetch" |
| remember | no tags | |
| related | untrusted-input | untrusted-input: keyword "fetch" |
| update_memory | no tags | |
| forget | no tags | |
| build_context | no tags |
Toxic-flow graph
The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.
The public result for this release does not print the flow graph, so there is nothing to show here. That is not the same as "no paths were found": what the scan did read is above, under coverage.
Supply chain and provenance
This is the first scan of this surface here, so there is nothing yet to compare it against.
Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.
Release history
Always latest — the install command names no version, so it installs the newest release when it runs; the current release on file is 0.16.1. Pin a release below and the command asks for exactly that one.
No release note was published with this version.
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Any signed-in account that does not publish Memoria can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.