Skip to content

Mentionry MCP server

com.mentionry/mentionrymentionry.comstreamable-httpv1.0.0Scanned

Find who links to your competitors but not you, judge each opening, and draft the pitch.

43 tools: 1 can delete or cancel things, 13 change things and 29 only read.

Tools: 1 delete, 0 irreversible, 13 write, 29 read.
  • Reads private data · 4
  • Reads text others write · 20
  • Can send data out · 15
  • 2 findings

Tools listed without sign-in. Remote posture

With Anlyon

1 tool here can delete or cancel things. Govern it.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

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 Mentionry MCP server can do

The Mentionry MCP server is published by mentionry.com, a domain the MCP Registry verified. It has 43 tools. It can delete or cancel things (1 tool: cancel_preparation_job) and create or change data (13 tools: review_answer_evidence, restore_answer_evidence, review_preparation_action, dismiss_preparation_action and 9 more). 29 tools only read.

4 tools read private data (find_prospects, list_documents, read_document and 1 more); 20 tools bring in text other people can write (find_prospects, read_visibility, read_latest_answer_reviews and 17 more); 15 tools can send data out of your system (write_email_template, read_search_console_performance, read_live_search_campaign and 12 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 2 findings (tool-shadowing and annotation-conflict). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.

Last scanned Nov 13, 2026, version 1.0.0.

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.

43 tools

Deletes or cancels 1Claude Code: deny
  • cancel_preparation_jobCancel a queued job.1 finding

    Why it is in this group

    • The name "cancel_preparation_job" says "cancel": a word for removing or ending something.
    • The server marks it not destructive (destructiveHint: false), but an annotation never lowers a class.

    Findings

    • Warn

      The server marks "cancel_preparation_job" not destructive (destructiveHint: false), and its own name, description or input give the class "destructive".

      Matched text

      annotations: {"readOnlyHint":false,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "cancel_preparation_job" says "cancel": a word for removing or ending something.

      What this check looks for
    Description, as the server wrote it
    Cancel a queued job. Running work can already have provider cost or a completed artifact; cancellation cannot guarantee a refund or rollback.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
Changes things 13Claude Code: ask
  • dismiss_preparation_actionDismiss suggested/reviewed/prepared work without deleting artifacts.No findings

    Why it is in this group

    • The name "dismiss_preparation_action" says "dismiss": a word for a change.
    Description, as the server wrote it
    Dismiss suggested/reviewed/prepared work without deleting artifacts. Requires expectedVersion and idempotencyKey. Prevents future automatic preparation; does not undo already running work or applied evidence.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • draft_pitchWrite one outreach email for one prospect, in your own words, using the reason that prospect would link to you.No findings

    Why it is in this group

    • The description's first verb is "write": a word for a change.
    Description, as the server wrote it
    Write one outreach email for one prospect, in your own words, using the reason that prospect would link to you. It returns the subject and body and sends nothing: sending happens from your own mailbox, after you have read it. Requires a paid Mentionry plan.
  • prepare_page_changesQueue a source-grounded draft for one saved mapped action.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Queue a source-grounded draft for one saved mapped action. May consume one draft and up to $0.02 planned provider spend inside the existing subscription budget. Requires an idempotency key. Does not publish or send.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • qualify_domainJudge whether a domain is worth running link building for at all, from live search data.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.
    Description, as the server wrote it
    Judge whether a domain is worth running link building for at all, from live search data. Returns a verdict, the measured signals behind it, and the honest case against. This is the same free diagnostic the dashboard runs, and for a large share of sites the answer is that the niche has nothing to mine.
  • record_page_deploymentDeclare that reviewed saved changes were applied to their original mapped page.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Declare that reviewed saved changes were applied to their original mapped page. Queues one bounded public HTML check, not CMS publishing, paid research or proof of ranking lift.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • restore_answer_evidenceRestore a reviewed recorded run to the review queue using stateVersion and idempotencyKey.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.
    Description, as the server wrote it
    Restore a reviewed recorded run to the review queue using stateVersion and idempotencyKey. No provider work, content publication or outcome claim.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • restore_preparation_actionRestore dismissed/archived work with expectedVersion and idempotencyKey.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.
    Description, as the server wrote it
    Restore dismissed/archived work with expectedVersion and idempotencyKey. Preserve existing applied/prepared evidence or return to suggested. No immediate provider work/job submission. Fresh suggestions can resume scheduled preparation; failed drafts remain explicit retries.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • review_answer_evidenceMark one recorded answer run reviewed, not published or improved.No findings

    Why it is in this group

    • The description's first verb is "mark": a word for a change.
    Description, as the server wrote it
    Mark one recorded answer run reviewed, not published or improved. Requires its saved stateVersion and idempotencyKey. New runs remain separate. No provider work.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • review_preparation_actionMark a suggestion reviewed using expectedVersion and idempotencyKey.No findings

    Why it is in this group

    • The description's first verb is "mark": a word for a change.
    Description, as the server wrote it
    Mark a suggestion reviewed using expectedVersion and idempotencyKey. Task triage only, not copy approval, generation, publication or outcome verification.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • run_agentRun one of your agents once and return what every step did, in order, with what it returned and how long it took.No findings

    Why it is in this group

    • The name "run_agent" says "run": a word for a change.
    Description, as the server wrote it
    Run one of your agents once and return what every step did, in order, with what it returned and how long it took. Pass its declared inputs as an object keyed by their `key` from describe_agent. Uses one included agent run. Provider calls remain subject to the shared collection allowance. Requires a paid Mentionry plan.
  • write_email_templateWrite a complete HTML marketing email (tables and inline styles, so it renders in Outlook and Gmail) from a brief and file it in Documents for preview.Can send data outNo findings

    Why it is in this group

    • The name "write_email_template" says "write": a word for a change.

    What it can reach

    • outboundThe name "write_email_template" says "write" and "email": it writes something other people read.
    Description, as the server wrote it
    Write a complete HTML marketing email (tables and inline styles, so it renders in Outlook and Gmail) from a brief and file it in Documents for preview. Grounded in the brand kit when one exists. The result can go into the customer's Gmail drafts with create_email_draft or be sent through an app action, both of which ask first. Spends one draft allowance.
  • write_landing_pageWrite a complete, responsive landing page as HTML from a brief and file it in Documents, where the customer previews it.No findings

    Why it is in this group

    • The name "write_landing_page" says "write": a word for a change.
    Description, as the server wrote it
    Write a complete, responsive landing page as HTML from a brief and file it in Documents, where the customer previews it. Grounded in the brand kit when one exists. Pass the addresses of pictures you generated so the page uses them. It is private until publish_page. Spends one draft allowance.
Only reads 29Show 29 read-only tools
  • describe_agentWhat one agent does, step by step, and what it needs: the values you must pass to run_agent, what it hands back, how many of its steps reach a vendor that charges, and anything cur…No findings

    Why it is in this group

    • The name "describe_agent" says "describe": a word for reading.
    Description, as the server wrote it
    What one agent does, step by step, and what it needs: the values you must pass to run_agent, what it hands back, how many of its steps reach a vendor that charges, and anything currently stopping it from running.
  • discover_competitorsFind the domains competing with this one, from both search results and link profile overlap.No findings

    Why it is in this group

    • The description's first verb is "find": a word for reading.
    Description, as the server wrote it
    Find the domains competing with this one, from both search results and link profile overlap. Returns each candidate with the evidence for it, so a wrong suggestion can be recognised rather than trusted.
  • find_prospectsSearch the open web for pages that would plausibly link to this site: resource pages, roundups, listicles and directories in its subject.Reads private dataReads text others writeNo findings

    Why it is in this group

    • The name "find_prospects" says "find": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "search" and "directory": it reads files.
    • untrusted-inputThe description's first sentence says "search" and "web": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Search the open web for pages that would plausibly link to this site: resource pages, roundups, listicles and directories in its subject. Returns each with the reason it was kept.
  • list_agentsThe agents on this account: what each is called, how many steps it has, whether it would run as it stands, and when it last changed.No findings

    Why it is in this group

    • The name "list_agents" says "list": a word for reading.
    Description, as the server wrote it
    The agents on this account: what each is called, how many steps it has, whether it would run as it stands, and when it last changed. Call describe_agent next to learn what one needs before running it.
  • list_brand_workspacesList your owned active brand workspaces for subsequent reads.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "list_brand_workspaces" says "list": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    List your owned active brand workspaces for subsequent reads.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • list_connectionsWhich outside tools this account has connected, and what each one is allowed to do here.No findings

    Why it is in this group

    • The name "list_connections" says "list": a word for reading.
    Description, as the server wrote it
    Which outside tools this account has connected, and what each one is allowed to do here. Call it before offering to file, post or draft anything, so you can say what is actually available instead of guessing.
  • list_documentsThe documents your agents have produced: what each is called, which agent wrote it, how long it is and how many versions it has.Reads private dataNo findings

    Why it is in this group

    • The name "list_documents" says "list": a word for reading.

    What it can reach

    • private-dataThe name "list_documents" says "list" and "document": it reads documents.
    • private-dataThe description's first sentence says "document": it reads documents.
    Description, as the server wrote it
    The documents your agents have produced: what each is called, which agent wrote it, how long it is and how many versions it has. Pass an id to read_document for the text.
  • list_mediaThe pictures and clips already generated on this account, newest first, with their addresses.No findings

    Why it is in this group

    • The name "list_media" says "list": a word for reading.
    Description, as the server wrote it
    The pictures and clips already generated on this account, newest first, with their addresses. Free.
  • read_actionable_insightsPrioritized page actions with evidence, implementation steps, confidence, limits and outcome verification.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_actionable_insights" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Prioritized page actions with evidence, implementation steps, confidence, limits and outcome verification.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_agent_capabilitiesDiscover saved resources, plan coverage, shared collection ceilings, protocol links and sample workflow without provider spending.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_agent_capabilities" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Discover saved resources, plan coverage, shared collection ceilings, protocol links and sample workflow without provider spending.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_brand_overviewSaved brand coverage and next SEO/GEO actions; no paid collection.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_brand_overview" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved brand coverage and next SEO/GEO actions; no paid collection.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_brand_resultsSaved 30-day answer-engine visibility and competitors, with explicit denominator and sample limits.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_brand_results" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved 30-day answer-engine visibility and competitors, with explicit denominator and sample limits.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_documentThe text of one document your agent wrote, with the agent, the step and the values that run was given.Reads private dataNo findings

    Why it is in this group

    • The name "read_document" says "read": a word for reading.

    What it can reach

    • private-dataThe name "read_document" says "read" and "document": it reads documents.
    • private-dataThe description's first sentence says "document": it reads documents.
    Description, as the server wrote it
    The text of one document your agent wrote, with the agent, the step and the values that run was given. Returns the newest version unless an older one is asked for.
  • read_latest_answer_reviewsLatest saved answer executions needing an owned-brand attribution review, with engine, recorded region/date, actual competing brands, excerpt, citations and a source-grounded revie…Reads text others writeNo findings

    Why it is in this group

    • The name "read_latest_answer_reviews" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe name "read_latest_answer_reviews" says "read" and "review": it reads comments, which other people can write.
    • untrusted-inputThe description's first sentence says "review": it reads comments, which other people can write.
    Description, as the server wrote it
    Latest saved answer executions needing an owned-brand attribution review, with engine, recorded region/date, actual competing brands, excerpt, citations and a source-grounded review task. Last seven days; at most 200 question/engine/region/persona groups scanned per brand. Up to 30 pending and 30 reviewed/unchanged runs with triage state/version. triage.unchangedSinceReview references a prior explicit review of complete identical evidence; the current run remains reviewed=false. Use triage.needsReview for fresh tasks; changed or unverified earlier reviews need rechecking; state changes use dedicated review/restore tools. No comparison, provider calls or publication.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_live_search_campaignSaved live Google desktop rank observations, full-text database suggestions and complete-link SERP-overlap groups.Reads private dataReads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_live_search_campaign" says "read": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "database": it reads databases.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved live Google desktop rank observations, full-text database suggestions and complete-link SERP-overlap groups. Includes dated compatible rank movement, explicit unknown gaps, leading-page changes and historical mapped-page associations (14 checks/90 days). URL switching is not proven cannibalization. Includes same-context sampled site overlap, not business-rival or complete-gap inference. No refresh or provider calls.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_page_connectionsActual sampled incoming/outgoing links with anchor text and nofollow evidence.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_page_connections" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Actual sampled incoming/outgoing links with anchor text and nofollow evidence.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_page_deploymentsRead saved deployment declarations and public HTML observations.Reads text others writeNo findings

    Why it is in this group

    • The name "read_page_deployments" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "read" and "html": it reads web pages, which anyone can write.
    Description, as the server wrote it
    Read saved deployment declarations and public HTML observations. Never fetches a page or starts work.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_page_insightsOne mapped page, keyword topics, SEO/GEO action plan and internal connections.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_page_insights" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    One mapped page, keyword topics, SEO/GEO action plan and internal connections.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_page_keywordsAll collected keyword rows for a mapped page, with lexical topic labels; paginated, not complete keyword discovery.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_page_keywords" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    All collected keyword rows for a mapped page, with lexical topic labels; paginated, not complete keyword discovery.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_preparation_actionRead one owned mapped action, its current stateVersion and cursor-paginated audit history.No findings

    Why it is in this group

    • The name "read_preparation_action" says "read": a word for reading.
    Description, as the server wrote it
    Read one owned mapped action, its current stateVersion and cursor-paginated audit history. Saved data only; no provider work.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_preparation_actionsSaved owned-page action IDs, source evidence and preparation states.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_preparation_actions" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved owned-page action IDs, source evidence and preparation states. Use an action ID with prepare_page_changes; no provider work is triggered by this read.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_preparation_jobRead a saved preparation job status; never schedules work or calls paid providers.No findings

    Why it is in this group

    • The name "read_preparation_job" says "read": a word for reading.
    Description, as the server wrote it
    Read a saved preparation job status; never schedules work or calls paid providers.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_prepared_page_changesSaved source-grounded copy changes and current eligibility.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_prepared_page_changes" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved source-grounded copy changes and current eligibility. Add a mapped URL for an eligible full artifact. Review-only historical proposals are withheld and replaced with an intent-review task. No generation, publishing or provider requests.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_search_console_performanceSaved finalized query/page/country/device rows and matching prior28day samples.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_search_console_performance" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved finalized query/page/country/device rows and matching prior28day samples. Plan-limited mapped pages; no refresh, provider call or causal lift claim.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_search_rank_historySaved keyword positions by snapshot, URL and market.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_search_rank_history" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "url": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Saved keyword positions by snapshot, URL and market. Indexed database samples, not a live rank campaign; absent observations are unknown.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • read_siteRead a domain and return what we think it is about: a description, its audience, and the keywords it should be found for.1 finding

    Why it is in this group

    • The name "read_site" says "read": a word for reading.

    Findings

    • Warn

      The description contains "every other tool", wording about other tools or servers.

      Matched text

      description: "...should be found for. This is the brief every other tool takes as input, so call it first when y..."

      What this check looks for
    Description, as the server wrote it
    Read a domain and return what we think it is about: a description, its audience, and the keywords it should be found for. This is the brief every other tool takes as input, so call it first when you do not already have one.
  • read_visibilityThe account's own answer-engine measurements over a window: how often each tracked brand appears in recorded answers (visibility), its share of all mentions (share of voice), its m…Reads text others writeNo findings

    Why it is in this group

    • The name "read_visibility" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe description's first sentence says "mention": it reads comments, which other people can write.
    Description, as the server wrote it
    The account's own answer-engine measurements over a window: how often each tracked brand appears in recorded answers (visibility), its share of all mentions (share of voice), its mean position, and the ranking against named competitors. Group by day for a trend, or by model, region, persona, prompt or topic to find where a brand is weak. This is the data behind the dashboard's Answer Engine Insights.
  • read_website_pagesPages and readiness in the entitled saved website map.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "read_website_pages" says "read": a word for reading.

    What it can reach

    • untrusted-inputThe name "read_website_pages" says "read" and "website": it reads web pages, which anyone can write.
    • untrusted-inputThe description's first sentence says "website": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Pages and readiness in the entitled saved website map.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false

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.

1200 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 4 private-data tools breaks every one.

Reads private data4
Reads files, mail, messages, databases, secrets, calendars or repositories.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__mentionry__find_prospects",
      "mcp__mentionry__list_documents",
      "mcp__mentionry__read_document",
      "mcp__mentionry__read_live_search_campaign"
    ]
  }
}

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.

Tools listed without sign-in

Looked at Nov 20, 2026, 18:42 UTC, unchanged since Oct 29, 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
14
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 14 tools that change things (1 destructive, 13 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://mentionry.com/api/mcp: tools/list returned 43 tools to an anonymous client, including cancel_preparation_job (destructive), review_answer_evidence (write), restore_answer_evidence (write), review_preparation_action (write), dismiss_preparation_action (write) and 9 more

    What this check looks for

Posture changes

  1. Oct 29, 2026
    • Tools that change things, listed without credentials: 12 to 14

Findings

Matches of fixed checks, with the text they matched.

2 warn
  • The description contains "every other tool", wording about other tools or servers.

    Matched text

    description: "...should be found for. This is the brief every other tool takes as input, so call it first when y..."

    What this check looks for
  • The server marks "cancel_preparation_job" not destructive (destructiveHint: false), and its own name, description or input give the class "destructive".

    Matched text

    annotations: {"readOnlyHint":false,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "cancel_preparation_job" says "cancel": a word for removing or ending something.

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

  1. Tools26 added.+26 addedView diff
  2. PostureRemote posture changed. Tools that change things, listed without credentials: 12 -> 14.View diff
    • Tools that change things, listed without credentials: 12 to 14

Install

Pinned to version 1.0.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http mentionry https://mentionry.com/api/mcp --header 'Authorization: <your-authorization>'
claude mcp add --transport http mentionry https://mentionry.com/api/mcp --header 'Authorization: <your-authorization>'

Run 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.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__mentionry__read_preparation_action",
      "mcp__mentionry__read_site",
      "mcp__mentionry__discover_competitors",
      "mcp__mentionry__backlink_gap",
      "mcp__mentionry__find_prospects",
      "mcp__mentionry__list_agents",
      "mcp__mentionry__describe_agent",
      "mcp__mentionry__list_documents",
      "mcp__mentionry__read_document",
      "mcp__mentionry__read_visibility",
      "mcp__mentionry__list_connections",
      "mcp__mentionry__list_media",
      "mcp__mentionry__read_latest_answer_reviews",
      "mcp__mentionry__read_page_deployments",
      "mcp__mentionry__read_preparation_job"
    ],
    "ask": [
      "mcp__mentionry__review_answer_evidence",
      "mcp__mentionry__restore_answer_evidence",
      "mcp__mentionry__review_preparation_action",
      "mcp__mentionry__dismiss_preparation_action",
      "mcp__mentionry__restore_preparation_action",
      "mcp__mentionry__qualify_domain",
      "mcp__mentionry__verify_links",
      "mcp__mentionry__draft_pitch",
      "mcp__mentionry__run_agent",
      "mcp__mentionry__write_landing_page",
      "mcp__mentionry__write_email_template",
      "mcp__mentionry__read_search_console_performance",
      "mcp__mentionry__read_live_search_campaign",
      "mcp__mentionry__read_preparation_actions",
      "mcp__mentionry__read_search_rank_history",
      "mcp__mentionry__read_prepared_page_changes",
      "mcp__mentionry__read_agent_capabilities",
      "mcp__mentionry__read_brand_overview",
      "mcp__mentionry__read_actionable_insights",
      "mcp__mentionry__read_website_pages",
      "mcp__mentionry__read_page_insights",
      "mcp__mentionry__read_page_keywords",
      "mcp__mentionry__read_page_connections",
      "mcp__mentionry__read_brand_results",
      "mcp__mentionry__list_brand_workspaces",
      "mcp__mentionry__record_page_deployment",
      "mcp__mentionry__prepare_page_changes"
    ],
    "deny": [
      "mcp__mentionry__cancel_preparation_job"
    ]
  }
}

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 mentionry. Review it before you use it.

Capability tags behind 23 rules
  • mcp__mentionry__find_prospects private-data, untrusted-input
  • mcp__mentionry__list_documents private-data
  • mcp__mentionry__read_document private-data
  • mcp__mentionry__read_visibility untrusted-input
  • mcp__mentionry__write_email_template outbound
  • mcp__mentionry__read_latest_answer_reviews untrusted-input
  • mcp__mentionry__read_search_console_performance untrusted-input, outbound
  • mcp__mentionry__read_live_search_campaign private-data, untrusted-input, outbound
  • mcp__mentionry__read_preparation_actions untrusted-input, outbound
  • mcp__mentionry__read_search_rank_history untrusted-input, outbound
  • mcp__mentionry__read_prepared_page_changes untrusted-input, outbound
  • mcp__mentionry__read_agent_capabilities untrusted-input, outbound
  • mcp__mentionry__read_brand_overview untrusted-input, outbound
  • mcp__mentionry__read_actionable_insights untrusted-input, outbound
  • mcp__mentionry__read_website_pages untrusted-input, outbound
  • mcp__mentionry__read_page_insights untrusted-input, outbound
  • mcp__mentionry__read_page_keywords untrusted-input, outbound
  • mcp__mentionry__read_page_connections untrusted-input, outbound
  • mcp__mentionry__read_brand_results untrusted-input, outbound
  • mcp__mentionry__list_brand_workspaces untrusted-input, outbound
  • mcp__mentionry__record_page_deployment untrusted-input
  • mcp__mentionry__read_page_deployments untrusted-input
  • mcp__mentionry__prepare_page_changes 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.mentionry/mentionry/verify?hash=0190e76308d7e5ac2e77603d32196b53633d2f798fdabff14dde8a7ac21c8100" | jq -e '.changed == false'

Pinned hash 0190e76308d7, from Nov 13, 2026, 19:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/com.mentionry/mentionry/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for com.mentionry/mentionry
[![Anlyon Radar](https://radar.anlyon.com/badge/com.mentionry/mentionry.svg)](https://radar.anlyon.com/servers/com.mentionry/mentionry)

Source: MCP Registry. Synced Oct 21, 2026, 17:59 UTC.

Mentionry MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the Mentionry MCP server safe?

No scan can promise that. What its definitions show: 43 tools, 14 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 Mentionry MCP server delete data?

Yes. 1 tool can delete, cancel or revoke something: cancel_preparation_job.

Can the Mentionry MCP server do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the Mentionry MCP server have?

43 tools: 1 destructive, 13 write, 29 read. cancel_preparation_job, review_answer_evidence, restore_answer_evidence, review_preparation_action, dismiss_preparation_action, restore_preparation_action, qualify_domain, verify_links and 35 more.

Does the Mentionry MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the Mentionry MCP server send data to outside services?

15 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): write_email_template, read_search_console_performance, read_live_search_campaign, read_preparation_actions, read_search_rank_history, read_prepared_page_changes and 9 more.

Has the Mentionry MCP server changed its tools?

Yes, once since Anlyon Radar started tracking it, most recently on Oct 31, 2026. Every change is on this page and in its Atom feed.