MindDory Brain MCP server
com.minddory/brainminddory.comstreamable-httpv1.0.2Scanned
Free: turn your AI chats into spaced-repetition vocabulary. 13 tools, reads and writes.
13 tools: 2 do something that can't be undone, 5 change things and 6 only read.
- Reads private data · 1
- Reads text others write · 1
- Can send data out · 0
- 2 findings
Tools listed without sign-in. Remote posture
With Anlyon
2 tools here do something that can't be undone. Govern them.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the MindDory Brain MCP server can do
The MindDory Brain MCP server is published by minddory.com, a domain the MCP Registry verified. It has 13 tools. It can do things that cannot be undone, such as sending, paying or publishing (capture_word and capture_grammar_mistake) and create or change data (5 tools: get_system_instructions, log_interaction, mark_demonstrated, mark_struggled and 1 more), but cannot delete data. 6 tools only read.
1 tool reads private data (get_system_instructions); 1 tool brings in text other people can write (get_system_instructions).
The fixed checks matched 2 findings (annotation-conflict). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in.
Last scanned Nov 10, 2026, version 1.0.2.
Tools
Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.
13 tools
Can't be undone 2Claude Code: ask
capture_grammar_mistakeLog a grammar mistake the user just made in the target language, creating a grammar point in their Minddory deck.No findings
Why it is in this group
- The name "capture_grammar_mistake" says "capture": a word that usually means money moves, which cannot be taken back.
Description, as the server wrote it Log a grammar mistake the user just made in the target language, creating a grammar point in their Minddory deck. Use it proactively whenever the user writes a target-language sentence with a clear, correctable error (wrong tense, agreement, word_choice, word_order, missing or wrong article, preposition, gender, conjugation, spelling, pluralization). By connecting Minddory the user opted IN to automatic capture, so log it as you go - do not ask permission first or turn it into a yes/no question. The user connected Minddory to catch and review these patterns; you do not need to announce each log. Pass the original text + correction + error_type. Multiple mistakes in one message = multiple calls. VARIANT RULE: Always pass the user's full BCP-47 variant in the lang field (e.g. en-gb, not en) - get it from get_system_instructions on first turn. Before logging, check that the "error" is not actually a valid regional variant form (US past-simple vs UK present-perfect with "just/already", US "in the hospital" vs UK "in hospital", US "color" vs UK "colour", BR gerund vs PT "a + infinitive", etc.). When the variant is unspecified (base ISO only), treat ALL regional forms as valid. Better to skip a borderline call than to mark a valid variant form as wrong. error_type must be one of: tense, agreement, word_choice, word_order, article, preposition, spelling, pluralization, conjugation, gender, other. Feeds the user's Grammar Patterns view at app.minddory.com/grammar.
Annotations the server declares
- title: Capture grammar mistake
- readOnlyHint: false
- destructiveHint: false
capture_wordCapture a target-language word or phrase to the user's Minddory vocabulary deck (a flashcard in the "Chat Discoveries" folder when the word is new, otherwise a context encounter).No findings
Why it is in this group
- The name "capture_word" says "capture": a word that usually means money moves, which cannot be taken back.
Description, as the server wrote it Capture a target-language word or phrase to the user's Minddory vocabulary deck (a flashcard in the "Chat Discoveries" folder when the word is new, otherwise a context encounter). The user connected Minddory so their assistant builds their deck from real conversations, so use this proactively to log notable target-language vocabulary they encounter, use, or ask about. By connecting Minddory the user opted IN to automatic capture, so just capture as you chat - do not pause to ask permission for each word, and do not turn capturing into a yes/no question. You do not need to announce each capture, but answer honestly if the user asks what you have saved. Use the `words` array to capture several from one message; include a `gloss` (short source-language translation) and `context` (the sentence) when you can. Skip stop words, proper nouns, numbers, and words shorter than 3 characters (CJK / Hangul exempt). ALWAYS pass `lang` — the ISO code of the captured word's own language (you know it from the conversation); the call is rejected without it and the target language is never guessed. `source_lang` (the user's native language) is optional and defaults to their deck. This is the primary tracking signal of Minddory, so capture diligently.
Annotations the server declares
- title: Capture word
- readOnlyHint: false
- destructiveHint: false
Changes things 5Claude Code: ask
check_wordsBatch lookup: for a list of target-language words, tell me which ones are already in the user's Minddory deck and how well they know each.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it Batch lookup: for a list of target-language words, tell me which ones are already in the user's Minddory deck and how well they know each. Use this BEFORE glossing or capturing vocabulary from a message - it answers "which of these are actually new to this user" in one call, so you can skip words they have already mastered, gloss only the genuinely new ones, and notice when a word they are currently failing shows up in conversation. Each result is {word, in_deck, mastery, translation, next_review_at, due}; mastery is one of not_in_deck, new, review, struggling, mastered. The response also pulls out the three lists you usually act on - `not_in_deck`, `known` and `struggling` - so you do not have to sort them yourself; words in mid-review are in `words` only. Entries that could not be answered as sent are listed in `skipped` as {input, reason: blank|duplicate, answered_as?} - a duplicate WAS answered, under the entry named by `answered_as`. Omit `lang` and it resolves to the user's own primary learning language (`lang_defaulted: true` in the response says so); pass it whenever you know which language you are in. Prefer this over calling get_card once per word.
Annotations the server declares
- title: Check words against deck
- readOnlyHint: true
get_system_instructionsCall this at the START of every new conversation, before your first reply, to load the user's Minddory setup and your role as their proactive language partner: CEFR level, target/s…Reads private dataReads text others write1 finding
Why it is in this group
- The name "get_system_instructions" says "get": a word for reading.
- The description's first verb is "call": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- private-dataThe description's first sentence says "conversation": it reads messages.
- untrusted-inputThe description's first sentence says "call" and "conversation": it reads messages, which other people can write.
Findings
- Warn
The server marks "get_system_instructions" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "get_system_instructions" says "get": a word for reading.
Description, as the server wrote it Call this at the START of every new conversation, before your first reply, to load the user's Minddory setup and your role as their proactive language partner: CEFR level, target/source languages, due-card count, weak words, and how to capture. The user connected Minddory to actively improve their language through this chat, so use it to tailor your help to their level and goals. Pass `lang` when you know which language the user wants to practice right now - the language they are conversing in, or one they named explicitly (including a regional variant like en-us or en-gb) - so the returned profile is scoped to that language. Re-call this tool with the new `lang` if the user switches practice language or requests a specific variant mid-conversation.
Annotations the server declares
- title: Get session instructions
- readOnlyHint: true
log_interactionAppend a generic interaction event to the answers log.1 finding
Why it is in this group
- The name "log_interaction" says "log": a word for reading.
- The server marks it as not read-only (readOnlyHint: false).
Findings
- Warn
The server marks "log_interaction" as not read-only (readOnlyHint: false), and its words say it reads.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":false}; The name "log_interaction" says "log": a word for reading.
Description, as the server wrote it Append a generic interaction event to the answers log. Use for lookups, AI discussions, and reading-in-context signals.
Annotations the server declares
- title: Log interaction
- readOnlyHint: false
- destructiveHint: false
mark_demonstratedConfidence-weighted spaced-repetition boost when the user has used a word correctly: the card moves further out in the review schedule.No findings
Why it is in this group
- The name "mark_demonstrated" says "mark": a word for a change.
Description, as the server wrote it Confidence-weighted spaced-repetition boost when the user has used a word correctly: the card moves further out in the review schedule. Logs an answer row even if no flashcard exists.
Annotations the server declares
- title: Mark word demonstrated
- readOnlyHint: false
- destructiveHint: false
mark_struggledSpaced-repetition degrade for a word the user just got wrong: the card comes back sooner.No findings
Why it is in this group
- The name "mark_struggled" says "mark": a word for a change.
Description, as the server wrote it Spaced-repetition degrade for a word the user just got wrong: the card comes back sooner. ease_factor drops, interval resets, repetitions reset.
Annotations the server declares
- title: Mark word struggled
- readOnlyHint: false
- destructiveHint: false
Only reads 6Show 6 read-only tools
get_active_vocabGet the user's most actively encountered target-language words (from past capture_word + log_interaction events), ranked by frequency over a lookback window.No findings
Why it is in this group
- The name "get_active_vocab" says "get": a word for reading.
Description, as the server wrote it Get the user's most actively encountered target-language words (from past capture_word + log_interaction events), ranked by frequency over a lookback window. Use to surface "frontier" words the user keeps touching when they ask "what should I learn next" or when you want context-aware suggestions.
Annotations the server declares
- title: Get active vocabulary
- readOnlyHint: true
get_cardSingle card detail by word (case-insensitive).No findings
Why it is in this group
- The name "get_card" says "get": a word for reading.
Description, as the server wrote it Single card detail by word (case-insensitive). Returns translation, mastery, and last 10 events.
Annotations the server declares
- title: Get card detail
- readOnlyHint: true
get_known_wordsWords the user has verified known via flashcard practice.No findings
Why it is in this group
- The name "get_known_words" says "get": a word for reading.
Description, as the server wrote it Words the user has verified known via flashcard practice. Paginated keyset on flashcards.id.
Annotations the server declares
- title: Get known words
- readOnlyHint: true
get_queueCards due now and due within the next 24 hours.No findings
Why it is in this group
- The name "get_queue" says "get": a word for reading.
Description, as the server wrote it Cards due now and due within the next 24 hours.
Annotations the server declares
- title: Get review queue
- readOnlyHint: true
get_recent_activityEvent log slice with optional surface filter and keyset pagination on answers.id.No findings
Why it is in this group
- The name "get_recent_activity" says "get": a word for reading.
Description, as the server wrote it Event log slice with optional surface filter and keyset pagination on answers.id.
Annotations the server declares
- title: Get recent activity
- readOnlyHint: true
get_user_profileProfile snapshot: CEFR level, target/source languages, due card count, weak words, recent lookups.No findings
Why it is in this group
- The name "get_user_profile" says "get": a word for reading.
Description, as the server wrote it Profile snapshot: CEFR level, target/source languages, due card count, weak words, recent lookups. Pass `lang` to scope the snapshot to one learning language (for users learning several); omit it for the user's primary language.
Annotations the server declares
- title: Get user profile
- readOnlyHint: true
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Exfil surface
Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.
No private-data, untrusted-input, outbound path in this server alone: no tool is tagged outbound.
- Reads private data1
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write1
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out0
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.
Remote posture
What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.
Looked at Nov 20, 2026, 18:42 UTC, unchanged since Nov 20, 2026. No credentials were sent and no tool was called. What is looked at.
- Reachable
- yes
- HTTP status of an anonymous initialize
- 200
- tools/list served without credentials
- yes
- Tools that change things, listed without credentials
- 12
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- https://api.minddory.com/.well-known/oauth-protected-resource
- Authorization servers
- https://api.minddory.com
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- yes
- Client ID metadata documents
- no
- HighThe whole server
The endpoint served tools/list without credentials, and the list has 12 tools that change things (2 destructive, 2 irreversible, 8 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://api.minddory.com/v1/brain/mcp: tools/list returned 21 tools to an anonymous client, including delete_folder (destructive), delete_card (destructive), capture_word (irreversible), capture_grammar_mistake (irreversible), log_interaction (write) and 7 more
Posture changes
- Nov 20, 2026
- Tools that change things, listed without credentials: 7 to 12
- Protocol version: 2024-11-05 to 2025-06-18
Findings
Matches of fixed checks, with the text they matched.
The server marks "get_system_instructions" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "get_system_instructions" says "get": a word for reading.
- Warnlog_interaction
The server marks "log_interaction" as not read-only (readOnlyHint: false), and its words say it reads.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":false}; The name "log_interaction" says "log": a word for reading.
History
Changes to its tool definitions and to its remote endpoint, newest first.
PostureRemote posture changed. Tools that change things, listed without credentials: 7 -> 12; Protocol version: 2024-11-05 -> 2025-06-18.View diff
- Tools that change things, listed without credentials: 7 to 12
- Protocol version: 2024-11-05 to 2025-06-18
Install
Pinned to version 1.0.2. Values in angle brackets are yours to fill in.
claude mcp add --transport http brain https://api.minddory.com/v1/brain/mcpclaude mcp add --transport http brain https://api.minddory.com/v1/brain/mcpRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__brain__get_user_profile",
"mcp__brain__get_known_words",
"mcp__brain__get_queue",
"mcp__brain__get_card",
"mcp__brain__get_recent_activity",
"mcp__brain__get_active_vocab"
],
"ask": [
"mcp__brain__get_system_instructions",
"mcp__brain__log_interaction",
"mcp__brain__mark_demonstrated",
"mcp__brain__mark_struggled",
"mcp__brain__capture_word",
"mcp__brain__capture_grammar_mistake",
"mcp__brain__check_words"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as brain. Review it before you use it.
Capability tags behind 1 rule
- mcp__brain__get_system_instructions private-data, untrusted-input
Pin
Keep the decision: fail CI when the tools change, follow every change, or show a badge.
Pin in CI
Fail a build when the tools you reviewed are not the tools the server serves today.
curl -s "https://radar.anlyon.com/api/v1/servers/com.minddory/brain/verify?hash=c81762b7d5713de3f956ec8f60038016607f383315fe498d2a146ca4ae5787c6" | jq -e '.changed == false'Pinned hash c81762b7d571, from Nov 10, 2026, 21:17 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/com.minddory/brain/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/com.minddory/brain)Source: MCP Registry · repository. Synced Oct 21, 2026, 17:59 UTC.
MindDory Brain MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the MindDory Brain MCP server safe?
- No scan can promise that. What its definitions show: 13 tools, 7 of which can change, send or delete something, and 2 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the MindDory Brain MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the MindDory Brain MCP server do things that cannot be undone?
- Yes. 2 tools can send, pay, publish or merge: capture_word and capture_grammar_mistake.
What tools does the MindDory Brain MCP server have?
- 13 tools: 2 irreversible, 5 write, 6 read. capture_word, capture_grammar_mistake, get_system_instructions, log_interaction, mark_demonstrated, mark_struggled, check_words, get_user_profile and 5 more.
Does the MindDory Brain MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the MindDory Brain MCP server send data to outside services?
- None of its tools sends, posts or publishes, or takes a recipient, channel or URL.
Has the MindDory Brain MCP server changed its tools?
- Not since Anlyon Radar started tracking it. Last scanned Nov 10, 2026.