AgentAvow Trust

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

8 tools

Its own safety labels

8 read-only

Memory it takes

About 1,600 tokens of your AI's memory, just for its tool descriptions

Largest answer

About 1.8 KB in our test

Speed

Average answer in our test: 114 ms

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, 10 Nov 2026.

Is AgentAvow Trust MCP server safe?

• None of the hidden-instruction wordings we know were found in its 8 tool descriptions.
• Its error messages showed no secrets or internal details.
• All 8 of its tools are labeled read-only.
• Overall it scored 90/100 on 10 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

8 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

  1 warning

  • !check_interaction_safety: Inputs with no description: interaction_type
  note
• !
  Follows MCP best practices (not scored yet)

  3 tips (not scored)

  • !scan_repo: Labeled read-only, but it has a delete-style "force" input. A read-only label can't promise much when the tool runs whatever it's sent; make the tool itself refuse changes (for example with a read-only database user), or drop the label. QAm
  • !scan_package: Labeled read-only, but it has a delete-style "force" input. A read-only label can't promise much when the tool runs whatever it's sent; make the tool itself refuse changes (for example with a read-only database user), or drop the label. QAm
  • !scan_mcp_server: Labeled read-only, but it has a delete-style "force" input. A read-only label can't promise much when the tool runs whatever it's sent; make the tool itself refuse changes (for example with a read-only database user), or drop the label. QAm
  note
• !
  Resources and prompts work

  1 warning (2 resources, 2 prompts)

  • !Listing resource templates failed: Error POSTing to endpoint: {"jsonrpc":"2.0","id":26,"error":{"code":-32601,"message":"Method not found","data":"resources/templates/list"}}
  • and 3 more
  note
• ✓
  ConnectsConnected in 195 ms
  pass
• ✓
  Ready for the newest MCP versionUses the latest MCP version (2026-07-28)
  pass
• ✓
  Lists its toolsFound 8 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
• ✓
  Tools work with good inputs

  All good

  • ✓verify_trust: Worked (83 ms)
  • ✓check_interaction_safety: Worked (74 ms)
  • ✓lookup_identity: Worked (289 ms)
  • ✓get_trust_badge: Worked (75 ms)
  • ✓about_agentavow: Worked (47 ms)
  • –scan_repo: Not run: labeled read-only, but it has a delete-style "force" input, so what it does depends on what it's sent. Skipped to be safe: if you own this server, test again with --owner
  • –scan_package: Not run: labeled read-only, but it has a delete-style "force" input, so what it does depends on what it's sent. Skipped to be safe: if you own this server, test again with --owner
  • –scan_mcp_server: Not run: labeled read-only, but it has a delete-style "force" input, so what it does depends on what it's sent. Skipped to be safe: if you own this server, test again with --owner
  pass
• ✓
  Bad inputs give a clean error

  All good

  • ✓verify_trust: Refused missing required input "entity_id" with a clear error
  • ✓verify_trust: Refused wrong type for "entity_id" with a clear error
  • ✓check_interaction_safety: Refused missing required input "target_entity_id" with a clear error
  • ✓check_interaction_safety: Refused wrong type for "target_entity_id" with a clear error
  • ✓lookup_identity: Refused missing required input "query" with a clear error
  • ✓lookup_identity: Refused wrong type for "query" with a clear error
  • ✓get_trust_badge: Refused missing required input "entity_id" with a clear error
  • ✓get_trust_badge: Refused wrong type for "entity_id" with a clear error
  pass
• ✓
  Tools work together

  5 chains working

  • ✓lookup_identity → verify_trust: verify_trust found what lookup_identity returned (entity_id = "14/100")
  • ✓lookup_identity → check_interaction_safety: check_interaction_safety found what lookup_identity returned (target_entity_id = "14/100")
  • ✓lookup_identity → get_trust_badge: get_trust_badge found what lookup_identity returned (entity_id = "14/100")
  • ✓about_agentavow → verify_trust: verify_trust found what about_agentavow returned (entity_id = "Ed25519/JWS-signed")
  • ✓about_agentavow → check_interaction_safety: check_interaction_safety found what about_agentavow returned (target_entity_id = "Ed25519/JWS-signed")
  pass
• ✓
  Fast enough

  All good

  • ✓Connecting took 195 ms
  • ✓Listing tools took 47 ms
  • ✓Average tool answer: 114 ms
  • ✓lookup_identity: Slowest tool: lookup_identity (289 ms)
  pass
• ✓
  No secrets leakedChecked 8 error messages: no secrets or internal details
  pass

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

• 10 Nov 202690

One test so far. New tests will show here.

Reviews

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