Query Markup Documents - On-device hybrid search for markdown files with BM25, vector search, and LLM reranking
mindin 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
Indexed search, schema-checked writes and session hooks for a markdown notes vault
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user mindin-mcp '--' uvx mindin
# 2 — ask your agent for something these tools do
# search, read_file, write_file and 20 more
mindin 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 mindin 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 uv (docs.astral.sh/uv): uvx 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 mindin-mcp '--' uvx mindinReconnect, or start a new session, and the tools appear in the model’s tool list.
Nothing is required, but it can take VAULT_PATH (Directory holding the vault. Set this or the four S3_* variables.), S3_ENDPOINT (S3 or MinIO endpoint URL holding the vault), S3_ACCESS_KEY (Access key for the vault bucket), S3_SECRET_KEY (Secret key for the vault bucket), S3_BUCKET (Bucket holding the vault), S3_PREFIX (Key prefix inside the bucket), S3_REGION (Region for the vault bucket), VAULT_CACHE_DIR (Where the SQLite index lives), VAULT_SCHEMA (Local schema file overriding the one in the vault), VAULT_STOP_HOOK (Set to off to disable the stop hook), VAULT_STOP_HOOK_COMMITS (Set to off to turn off the stop hook's unlogged-commit and stale-note checks alone), VAULT_STOP_HOOK_PROMPTS (Messages the user writes after the agent's last vault write before the stop hook asks it to save them (default 3), or off) if you want to set them.
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.
23 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.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| query | string | yes | — |
| limit | integer | no | — |
| folder | string | no | — |
| status | string | no | — |
| tag | string | no | — |
| kind | string | no | — |
| area | string | no | — |
| include_archive | boolean | no | — |
| include_superseded | boolean | no | — |
| path_prefix | string | no | — |
| since | string | no | — |
MODERATE — one round trip to storage. Returns the note with an etag: first line; pass that etag back to write_file to refuse a write over someone else's change. Search first. .vault/schema.yml reads the vault's schema.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
WRITE — validates the frontmatter against the vault schema and refuses the write if it does not hold. Stamps updated with today and fills date when missing. Pass expected_etag from read_file to make the write conditional, or "*" to create a note that must not exist yet.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
| content | string | yes | — |
| expected_etag | string | no | — |
WRITE — appends to an existing note and bumps updated; the note must still satisfy the schema. A missing note is refused: create it with write_file.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
| content | string | yes | — |
WRITE — moves or renames a note; the note must satisfy the schema at the destination. Use close to archive finished work.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| source | string | yes | — |
| dest | string | yes | — |
WRITE — deletes a note for good. Prefer close, which keeps it.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
CHEAP — paths only, no metadata and no bodies. search answers context questions better.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| prefix | string | no | — |
WRITE — changes a note's status without rewriting it: backlog (with a priority: urgent, high, medium, low) parks it, active picks it up again, draft. The triage list at session start is settled with this, append_file or close.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
| status | string | yes | — |
| priority | string | no | — |
| source | string | no | — |
WRITE — finishes a note: sets status complete (or superseded with superseded_by when merged_into is given) and moves it into the archive. Run it in the same pass as the last commit of the work; a note left open is what makes the vault drift.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
| merged_into | string | no | — |
| status | string | no | — |
WRITE — files an idea as a backlog note in the task folder: title, area (stem or [[stem]]), one line saying what and why, optional priority (urgent, high, medium, low) and source (who said it and when, or a sha). Call it the moment something is deferred, in so many words or in passing; picking the…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| title | string | yes | — |
| area | string | yes | — |
| line | string | yes | — |
| priority | string | no | — |
| source | string | no | — |
CHEAP — the backlog sorted by priority then age. Filter by area (a stem, or the prefix a family of stems shares) and by priority.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| area | string | no | — |
| priority | string | no | — |
| limit | integer | no | — |
WRITE — appends one dated line to the repo log, with the commits it produced. Run once at the end of a session per repo. Creates the log note when it is missing. repo is the main checkout's name, also from a worktree; branch is the branch the commits were made on. area is the stem of the system…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| repo | string | yes | — |
| line | string | yes | — |
| commits | array | no | — |
| area | string | no | — |
| branch | string | no | — |
CHEAP — the session-start call. Returns the system notes covering the code at path, the open tasks for it, the reference notes and the tail of the repo log, in one answer. Use it before searching blind. format "json" adds the system notes' paths and dates and the triage keys, for hooks that read…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | no | — |
| repo | string | no | — |
| query | string | no | — |
| limit | integer | no | — |
| format | string | no | — |
WRITE — the weekly lint for a scheduled job: parks stale open task notes in the backlog when park is set, and writes the findings as the note at write when given; answers the findings.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| write | string | no | — |
| park | boolean | no | — |
WRITE — for tools that run agents on task notes, such as kebbie: writes which machine and task work on the note (frontmatter key kebbie), unless another machine holds it; force takes it over. An agent never calls it on its own. Answers JSON.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | yes | — |
| machine | string | yes | — |
| task | string | yes | — |
| repo | string | no | — |
| force | boolean | no | — |
WRITE — for tools that run agents on task notes, such as kebbie: clears the note's claim (frontmatter key kebbie), only when machine holds it if machine is given; without a path, clears every claim of machine and task. Answers JSON.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| path | string | no | — |
| machine | string | no | — |
| task | string | no | — |
CHEAP — for tools that draw a board from the vault, such as kebbie: every system and task note outside the archive and log folders with its metadata and claim, the opening paragraph of open tasks and the last lines of each repo log, under an etag. Pass since=<etag> to get only {"unchanged": true}…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| since | string | no | — |
CHEAP — for the command line's human output: the data behind search, backlog or context (command), with the arguments that command takes. Answers JSON.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| command | string | yes | — |
| query | string | no | — |
| limit | integer | no | — |
| area | string | no | — |
| priority | string | no | — |
| path | string | no | — |
| repo | string | no | — |
CHEAP — a repo's period page: its log lines, its commits and the notes dated in the period (YYYY-MM or YYYY). commits maps a day (YYYY-MM-DD) to its "sha subject" lines, read from the repo's git log by the caller; leave it out to show no commits. write stores the page as the period note instead of…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| repo | string | yes | — |
| period | string | yes | — |
| commits | object | no | — |
| write | boolean | no | — |
CHEAP — who this connection signs in as and whether it may write. Answers JSON.
Takes no parameters.
MODERATE — reads every note to report drift: broken frontmatter, missing required fields, missing area, unresolved wikilinks, orphans, stale open tasks, archived notes still marked active, duplicate stems and superseded targets that do not exist.
Takes no parameters.
WRITE — sets a vault up: writes .vault/schema.yml and the Obsidian Bases views into the vault root. Run once per vault, right after installing. preset picks the layout of a vault without a schema: …; without it, the vault gets default. On a vault that has a schema, it only writes the views that…
| Parameter | Type | Required | Notes |
|---|---|---|---|
| force | boolean | no | — |
| preset | string | no | — |
WRITE — mints the personal token Obsidian's Remotely Save plugin uses to reach this vault over WebDAV, and shows it once. Issuing a new token revokes the previous one. The username is the email on your identity.
Takes no parameters.
- 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 mindin 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 (23 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>".
16 of the 23 recovered tools are registered only when isinstance(resolver, Unconfigured) is false (mindin/server.py:86) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes none of the recovered tools.
6 of the 23 recovered tools are registered only when machine_tools holds (mindin/server.py:305) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes none of the recovered tools.
1 of the 23 recovered tools are registered only when personal_tokens is not None holds (mindin/server.py:461) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes none of the recovered tools.
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.
No tool is registered unconditionally · (1) 16 only when isinstance(resolver, Unconfigured) is false (mindin/server.py:86) · (2) 6 only when machine_tools holds (mindin/server.py:305) · (3) 1 only when personal_tokens is not None holds (mindin/server.py:461) — typically when a key or option is configured. All 23 are in the table below and were analysed; the conditional ones are marked.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| search conditional (1) | no tags | |
| delete_file conditional (1) | file-write | file-write: keyword "delete_file" in tool name |
| list_files conditional (1) | sensitive-source | sensitive-source: keyword "list_files" |
| whoami conditional (1) | no tags | |
| read_file conditional (1) | sensitive-source | sensitive-source: keyword "read_file" |
| write_file conditional (1) | file-write | file-write: keyword "write_file" in tool name |
| append_file conditional (1) | file-write | file-write: keyword "append_file" in tool name |
| move_file conditional (1) | file-write | file-write: keyword "move_file" in tool name |
| set_status conditional (1) | no tags | |
| close conditional (1) | no tags | |
| backlog_add conditional (1) | no tags | |
| backlog conditional (1) | no tags | |
| log_append conditional (1) | no tags | |
| context conditional (1) | no tags | |
| lint_and_park conditional (2) | no tags | |
| claim conditional (2) | no tags | |
| release conditional (2) | no tags | |
| snapshot conditional (2) | no tags | |
| view conditional (2) | code-exec | code-exec: parameter "command" |
| changelog conditional (2) | no tags | |
| lint conditional (1) | no tags | |
| init conditional (1) | no tags | |
| obsidian_access conditional (3) | 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.
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 0.15.0. Pin a release below and the command asks for exactly that one.
No release note was published with this version.
No release note was published with this version.
No release note was published with this version.
- The version this listing was on when mcprush began following its releases. The registry was not asked when it shipped, so this row carries no date.
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 mindin 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.