Sealmark
Automated re-test on 2026-11-20: not checked by hand, and not yet reviewed by the maintainers.

Results / tares

tares

version tested 1.60.0PyPItested 2026-11-20Sealmark 1.0.1methodology 0.3-draftSTANDARD run83 tools

Sealmark's automated test of tares 1.60.0 on 2026-11-20 gave it grade A, its lowest test-area grade (protocol A). None of the 83 tools could be called with default settings, so this grade covers the handshake and tool metadata only. Of its 10 checks, 4 passed, 2 inconclusive and 4 not applicable. 3 more checks are advisory, and advisory checks never change a grade: 3 not applicable. The run used Sealmark 1.0.1 with the standard settings.

A

Grade A

The server's grade is its lowest test-area grade. How grading works

None of the 83 tools could be called with default settings, so this grade covers the handshake and tool metadata only.

Grade by test area

A
Protocol4/4 points

Protocol checks

CheckVerdictWhat happened
protocol
Valid MCP handshake and tool schemas
pass
handshake ok, 83 well-formed tool(s)
Evidence for protocol: 1 line
  1. Probe initialize + tools/list
    Outcome server tares 1.30.0 answered; 83 tool(s)
    Decided from the MCP exchange (protocol)
tool-poisoningcritical
No tool-poisoning indicators in tool metadata
pass
scanned 83 tool definition(s) and serverInfo, including parameter text: nothing suspicious
robust-input
Survives malformed tool input
pass
91 invalid call(s) rejected cleanly
Evidence for robust-input: 1 line
  1. Probe 91 invalid call(s) across 19 tool(s)
    Outcome all rejected cleanly and the server stayed responsive
    Decided from the MCP exchange (protocol)
secret-echocritical
No environment secret echoed in responses
pass
canary secrets absent from tools/list and 91 captured response(s), 0 of them successful result(s)
instructions-hygiene (advisory)
Server instructions name only tools that exist
n/a
the server sent no instructions
refusal-quality (advisory)
Refusals say what to do next
n/a
no refusal was recorded in this run, so there is nothing to judge

Network checks

CheckVerdictWhat happened
net-loopback-ssrfcritical
Refuses loopback targets (SSRF)
inconclusive
the listener saw nothing, but 18 of 18 probes were not explicitly refused (18 failed without saying the target was refused), so a refusal cannot be told from a tool that cannot fetch at all (a reply that only reports a failed fetch (proxy, socket, DNS) is not read as a refusal; set net.control_url to a URL the tool can reach)
net-scheme-allowlistcritical
Only http(s) schemes accepted
inconclusive
no scheme had an observable effect, but 12 of 12 scored probes were not explicitly refused, so a refusal cannot be told from a tool that cannot fetch at all (a reply that only reports a failed fetch (proxy, socket, DNS) is not read as a refusal; set net.control_url to a URL the tool can reach)
net-caps
Response size and time are capped
n/a
net-loopback-ssrf was inconclusive: the server never reached the loopback listener, so no loopback URL is known to be fetchable; response size and time caps were not measured
net-redirect
Redirects are re-validated
n/a
the server refused or never reached the loopback listener, so redirect handling could not be observed (a server that blocks loopback cannot be sent a redirect it would follow)

Effects checks

CheckVerdictWhat happened
annotation-honestycritical
Read-only tools change nothing
n/a
no tool declares readOnlyHint: true
openworld-honesty
Closed-world tools stay on the machine
n/a
no tool declares openWorldHint: false that could be given a harness-owned control argument
undeclared-writes (advisory)
Unannotated tools leave the sandbox unchanged
n/a
no unannotated read-style tool could be given a harness-owned control argument

Each check id links to what it tests and how to fix a failure. The same text is in sealmark explain <id>.

Under a check, Evidence shows what it probed (sent or inspected) and what it observed in this run: up to 6 lines per check, each cut at 240 characters, for checks that are not security checks. The text is shortened, and paths and anything that looks like a secret or token are replaced. Run sealmark run --format json yourself for every line in full.

By tool

Which checks called each tool with a valid call in this run, and which did not exercise it. A check can leave a tool out for several reasons: a limit Sealmark sets, a tool left out by design, input the run could not fill in, or a call that returned an error. The CLI report gives the reason. A verdict belongs to the whole check, not to one tool: a failed check that called three tools does not say which one failed. Security checks are left out of this table.

ToolCalled byNot exercised by
readNo check listed here called it with a valid call.None
create_triggerNo check listed here called it with a valid call.None
update_triggerNo check listed here called it with a valid call.None
subscribeNo check listed here called it with a valid call.
catalog_listNo check listed here called it with a valid call.
catalog_describeNo check listed here called it with a valid call.None
rememberNo check listed here called it with a valid call.None
set_session_flowNo check listed here called it with a valid call.None
list_connectorsNo check listed here called it with a valid call.
discover_sourceNo check listed here called it with a valid call.None
discover_dockerNo check listed here called it with a valid call.
test_sourceNo check listed here called it with a valid call.None
create_sourceNo check listed here called it with a valid call.None
list_sourcesNo check listed here called it with a valid call.
source_fieldsNo check listed here called it with a valid call.None
list_projectsNo check listed here called it with a valid call.
join_projectNo check listed here called it with a valid call.
statsNo check listed here called it with a valid call.None
list_skillsNo check listed here called it with a valid call.None
get_skillNo check listed here called it with a valid call.None
list_findingsNo check listed here called it with a valid call.None
record_findingNo check listed here called it with a valid call.None
project_timelineNo check listed here called it with a valid call.None
create_projectNo check listed here called it with a valid call.None
list_peopleNo check listed here called it with a valid call.
write_docNo check listed here called it with a valid call.
list_docsNo check listed here called it with a valid call.
get_docNo check listed here called it with a valid call.
read_global_docsNo check listed here called it with a valid call.
add_to_globalNo check listed here called it with a valid call.
pick_upNo check listed here called it with a valid call.
list_ticketsNo check listed here called it with a valid call.
get_ticketNo check listed here called it with a valid call.
write_ticketNo check listed here called it with a valid call.
attach_working_docNo check listed here called it with a valid call.
link_ticketsNo check listed here called it with a valid call.
write_milestoneNo check listed here called it with a valid call.
record_milestone_checkNo check listed here called it with a valid call.
list_milestonesNo check listed here called it with a valid call.
get_integrationsNo check listed here called it with a valid call.
merge_readyNo check listed here called it with a valid call.
release_readyNo check listed here called it with a valid call.
set_required_checksNo check listed here called it with a valid call.
assign_ticketNo check listed here called it with a valid call.
set_stageNo check listed here called it with a valid call.
assumeNo check listed here called it with a valid call.
list_assumptionsNo check listed here called it with a valid call.
decide_assumptionNo check listed here called it with a valid call.
add_checkNo check listed here called it with a valid call.
remember_for_projectNo check listed here called it with a valid call.
recallNo check listed here called it with a valid call.
record_reviewNo check listed here called it with a valid call.
review_queueNo check listed here called it with a valid call.
review_briefNo check listed here called it with a valid call.
builder_briefNo check listed here called it with a valid call.
write_repo_mapNo check listed here called it with a valid call.
recurring_findingsNo check listed here called it with a valid call.
desk_addNo check listed here called it with a valid call.
desk_listNo check listed here called it with a valid call.
desk_answerNo check listed here called it with a valid call.
desk_withdrawNo check listed here called it with a valid call.
list_decisionsNo check listed here called it with a valid call.
get_crew_settingsNo check listed here called it with a valid call.
set_crew_settingsNo check listed here called it with a valid call.
list_secretsNo check listed here called it with a valid call.
read_grantsNo check listed here called it with a valid call.
add_grantNo check listed here called it with a valid call.
list_crewNo check listed here called it with a valid call.
get_stationNo check listed here called it with a valid call.
hand_overNo check listed here called it with a valid call.
set_laterNo check listed here called it with a valid call.
my_workNo check listed here called it with a valid call.
crew_projectsNo check listed here called it with a valid call.
release_projectNo check listed here called it with a valid call.
finish_projectNo check listed here called it with a valid call.
what_needs_meNo check listed here called it with a valid call.
how_is_projectNo check listed here called it with a valid call.
start_projectNo check listed here called it with a valid call.
hand_project_to_crewNo check listed here called it with a valid call.
what_is_brokenNo check listed here called it with a valid call.
how_is_serviceNo check listed here called it with a valid call.
what_happenedNo check listed here called it with a valid call.
mark_incidentNo check listed here called it with a valid call.

What was not covered

Grade over time

Since the run on

  • Server version changed from 1.59.0 to 1.60.0.
  • Input schema changed (any part of it, parameter wording included): get_doc, list_decisions, list_docs, list_milestones, list_tickets, my_work, review_brief, review_queue.

Unchanged: the same grade in every test area, the same Sealmark version (1.0.1).

  1. –did not run
  2. –did not run
  3. –did not run
  4. Av1.57.0Sealmark 1.0.1
  5. Av1.59.0Sealmark 1.0.1
  6. Av1.60.0Sealmark 1.0.1

Package facts

License
MIT AND LicenseRef-Tares-EE
Latest release
1.60.0, published 20 Nov 2026 (the version graded on this page)
Repository
glassflow/tares
Last push
20 Nov 2026
Open issues and pull requests
5

From PyPI and GitHub, checked 22 Nov 2026. Plain facts about the package; they play no part in the grade.

Add it to an MCP client

This starts version 1.60.0, the version graded on this page, with the start command from the test run. For a client that reads an mcpServers JSON file:

{
  "mcpServers": {
    "tares": {
      "command": "uvx",
      "args": [
        "tares==1.60.0",
        "mcp",
        "--transport",
        "stdio"
      ]
    }
  }
}

Or add it to Claude Code:

claude mcp add tares -- uvx tares==1.60.0 mcp --transport stdio

The test run started the server with no API keys or other secrets, and with a cleaned environment. If the server needs some, its own documentation says how to pass them.

Run it yourself

This repeats the run with the same Sealmark version and the standard settings, starting the server the same way:

npx -y sealmark@1.0.1 run --start "uvx tares==1.60.0 mcp --transport stdio"

To confirm your run used the standard settings, add --format json --out report.json to the command. In the report, config.standard should be true and config.hash should start with sha256:d13e7893a0ac, as they did in this run. If either differs, so did the settings, and the verdicts may too (STANDARD and CUSTOM runs).

To keep the setup, put the start command in sealmark.yaml as the server.

Grade alerts

One email when the grade of tares changes after a re-test. Nothing else, and every email has a link to stop them.

Or follow it in a feed reader: the Atom feed lists each move of the overall grade that came with a different version of this server, tested with the same Sealmark version as the graded run before.

Each of 613 verdicts was checked independently, and Sealmark was wrong on some: how reliable Sealmark is.

Is this wrong?

Reply on the GitHub issue or email disclosure@sealmark.dev and we will recheck it. Maintainers can ask for their server to be removed.

If we were wrong, we correct the page and add your reply. Security findings go to the maintainer privately first; see the disclosure policy.