LocalMark Hosted Agents

At a glance

Runs where

On their server. What you send to it goes to them.

What you need

Nothing: it answered without a login

Tools

5 tools

Its own safety labels

0 read-only, 5 labeled as changing things

Memory it takes

About 1,100 tokens of your AI's memory, just for its tool descriptions

Tokens are an estimate: the characters in its tool names, descriptions and inputs, divided by 4. The exact number depends on your AI model. Safety labels are what the server says about itself. From our latest test, 6 Nov 2026.

Is LocalMark Hosted Agents MCP server safe?

• None of the hidden-instruction wordings we know were found in its 5 tool descriptions.
• 0 of its 5 tools are labeled read-only; the others may change or send things.
• Overall it scored 90/100 on 6 Nov 2026.

A test can only see what the server shows from outside, not what it does with your data behind the scenes. Connect servers from makers you trust.

What this test does not check

• Its source code and packages.
• What a tool quietly does behind the scenes, such as writing a file or calling another service.
• Instructions hidden inside live tool answers. We only look for those in tool descriptions.
• Hidden instructions worded in a way our list doesn't know yet.
• Tools that could change something. Those are read, never run.

What QAmcp can and can't see · How to tell if an MCP server is safe · Hidden instructions · Prompt injection

Known security advisories

It lists no public code, so there is nothing to look up in GitHub Security Advisories or OSV.dev. QAmcp tests how a server behaves. It does not scan source code.

What it can do

5 tools, in the server's own words.

Test results

Safe mode: only tools labeled read-only were run. How the test works

• !
  Every tool is described

  5 warnings

  • !analyze_document: Inputs with no description: payload, allow_model_processing, request_id, payment_signature, force_new_task, skip_free_trial
  • !analyze_data: Inputs with no description: payload, allow_model_processing, request_id, payment_signature, force_new_task, skip_free_trial
  • !localize_content: Inputs with no description: payload, allow_model_processing, request_id, payment_signature, force_new_task, skip_free_trial
  • !advise_agentops: Inputs with no description: payload, allow_model_processing, request_id, payment_signature, force_new_task, skip_free_trial
  • !analyze_payment_offers: Inputs with no description: payload, allow_model_processing, request_id, payment_signature, force_new_task, skip_free_trial
  note
• ✓
  ConnectsConnected in 107 ms
  pass
• ✓
  Ready for the newest MCP versionUses the latest MCP version (2026-07-28)
  pass
• ✓
  Lists its toolsFound 5 tools
  pass
• ✓
  Tools are labeled safe or dangerousEvery tool has a safety label
  pass
• ✓
  No known hidden instructionsNone of the hidden-instruction wordings we know were found in tool descriptions. New wordings can get past this check.
  pass
• ✓
  Fast enough

  All good

  • ✓Connecting took 107 ms
  • ✓Listing tools took 27 ms
  pass
• ✓
  Follows MCP best practices (not scored yet)Tool names, safety labels and tool list size follow the MCP guide
  pass
• –
  Tools work with good inputs

  Not tested

  • –analyze_document: Not run: labeled as changing data
  • –analyze_data: Not run: labeled as changing data
  • –localize_content: Not run: labeled as changing data
  • –advise_agentops: Not run: labeled as changing data
  • –analyze_payment_offers: Not run: labeled as changing data
  not tested
• –
  Bad inputs give a clean errorNot tested
  not tested
• –
  Tools work togetherNo safe tool returned data to start a chain from, so chains were not tested (not counted either way)
  not tested
• –
  No secrets leakedNo error messages to check
  not tested

Test it yourself

Free and open source. Same checks, same score.

Show this score on GitHub

Free for every server in the directory. Paste this into your README: the badge shows the score from our latest test.

Score over time

• 6 Nov 202690

One test so far. New tests will show here.

Reviews

Coming soon: reviews from people who use LocalMark Hosted Agents, signed in with GitHub, one per person. Reviews never change the score.