<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p>Hi. Tim, Carlo and all,</p>
<p>From the side of China-VO, we have been put several MCP servers
and Agent Skills into operation since 2024. Please visit the
following links:</p>
<ul>
<li><a class="moz-txt-link-freetext" href="https://nadc.china-vo.org/agent-platform/">https://nadc.china-vo.org/agent-platform/</a></li>
<li><a class="moz-txt-link-freetext" href="https://nadc.china-vo.org/ai/chat/nadc_agent_ep">https://nadc.china-vo.org/ai/chat/nadc_agent_ep</a></li>
<li><a class="moz-txt-link-freetext" href="https://nadc.china-vo.org/ai/cms/article/view?id=10&locale=en">https://nadc.china-vo.org/ai/cms/article/view?id=10&locale=en</a></li>
</ul>
<p>For more information, please refer to my talk at the 2026 Spring
Interop meeting:</p>
<ul>
<li><a class="moz-txt-link-freetext" href="https://wiki.ivoa.net/internal/IVOA/InterOpJun2026Apps/nadc-ai4s-framework-20260610.pptx">https://wiki.ivoa.net/internal/IVOA/InterOpJun2026Apps/nadc-ai4s-framework-20260610.pptx</a></li>
</ul>
<p>We are very interested in defining related specifications on the
topic on behalf of the IVOA. A draft document will be available in
the coming months.</p>
<p>Cheers,</p>
<p>Chenzhou</p>
<p><br>
</p>
<div class="moz-cite-prefix">On 7/28/2026 9:07 PM, czwolf via
interop wrote:<br>
</div>
<blockquote type="cite"
cite="mid:26C5A576-5A8C-42C3-B424-53FF57C23D8E@obspm.fr">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
Hi Tim, all,
<div><br>
</div>
<div>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">During
the summer 2025 I set up a public MCP server (implementing
the Streamable http transport) to expose a SLAP2 data access
service in a way that LLM could extract and manipulate
data. </span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">It
worked just fine, but was not usable: the result from the
MCP response directly goes in the context window. Even
flagship models have a context of 1Mb and any “real” query
produced so many data that the context was immediately
saturated. </span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">We
adopted then the “skill” approach (this is what you describe
as the markdown giving instructions to the LLM). Check the
skill folder of <a href="https://github.com/VAMDC/pyVAMDC"
moz-do-not-send="true" class="moz-txt-link-freetext">https://github.com/VAMDC/pyVAMDC</a> to
see what we made. This approach works perfectly, and we are
very happy with this, but it requires a sandbox so that the
LLM can run system command and download files to parse. This
is the case for Claude, ChatGPT and we also achieved this
locally using open-wighted models + OpenWebUI with the
OpenTerminal extension.</span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">The
very true force of LLM is that they can adapt even if the
vocabulary of each data provider is slightly different. They
can cope with the differences and align the sense - they do
not need necessarily a fixed vocabulary. But this can help. </span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">I
recently participated in writing a recommendation from the
Research Data Alliance where we tried to define some
“standard” for agentic tool for research - maybe this could
be useful for the IVOA community too : <a
href="https://www.rd-alliance.org/groups/data-director-agentic-ai-blueprint/outputs/data-director-agentic-ai-tool-blueprint/"
moz-do-not-send="true" class="moz-txt-link-freetext">https://www.rd-alliance.org/groups/data-director-agentic-ai-blueprint/outputs/data-director-agentic-ai-tool-blueprint/</a></span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">All
the best, </span></p>
<p
style="margin: 0px 0px 18px; font-style: normal; font-variant-caps: normal; font-width: normal; line-height: normal; font-size-adjust: none; font-kerning: auto; font-variant-alternates: normal; font-variant-ligatures: normal; font-variant-numeric: normal; font-variant-east-asian: normal; font-variant-position: normal; font-feature-settings: normal; font-optical-sizing: auto; font-variation-settings: normal;"><span
style="font-kerning: none; background-color: rgba(255, 255, 255, 0);">Carlo.</span></p>
<div><span style="font-kerning: none"><br>
</span></div>
<div><br>
<blockquote type="cite">
<div>On 27 Jul 2026, at 17:09, Tim Jenness via interop
<a class="moz-txt-link-rfc2396E" href="mailto:interop@ivoa.net"><interop@ivoa.net></a> wrote:</div>
<br class="Apple-interchange-newline">
<div>
<div>
<div dir="ltr">At the recent SPIE meeting in Copenhagen,
Ashley Barnes from ESO gave a talk on their
experiments with teaching an LLM to query their data
archive. They looked at making a dedicated agent
taught explicitly but this cost them real money, and
they looked at writing markdown instructions that the
LLM could read to learn about the specifics of the
archive center. This led me to ponder whether we
should be talking about some kind of standardized MCP
server interface (<font color="#419cff"><span
style="-webkit-text-fill-color: rgb(65, 156, 255) !important;"><a
href="https://modelcontextprotocol.io/docs/getting-started/intro"
moz-do-not-send="true"
class="moz-txt-link-freetext">https://modelcontextprotocol.io/docs/getting-started/intro</a></span></font>)
that an LLM could query to work things out about the
archive.
<div><br>
</div>
<div dir="ltr">This seems to be the way that many
services are heading (my photos application has an
MCP server to let an LLM look for photos
efficiently). On the one hand we have all these open
protocols so that in theory an LLM can work it all
out and form the right query for a TAP server by
querying registry and looking at TAP_SCHEMA, and
that should work. MCP <i>might</i> be able to let
an agent do the same thing but using fewer tokens,
even if the end point is the agent sending off a TAP
query.</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">Is anyone else thinking about this? Can
someone from ESO ask Ashley to join IVOA Slack to
discuss this?</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">Each data center could put their own
MCP server up with their own targeted API and that
might be fine, but pooling our experience in terms
of what helped and what made things worse would be
really helpful and if we end up with an IVOA Note
containing advice on natural language queries that
would be great. If we ended up with a standardized
vocabulary for MCP servers that might be even
better.</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">-- </div>
<div dir="ltr">Tim Jenness</div>
<div dir="ltr">Rubin Observatory</div>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</blockquote>
<pre class="moz-signature" cols="72">--
===============================================================
Chenzhou Cui
National Astronomical Observatory | Tel: +86-10-64872500
Chinese Academy of Sciences | FAX: +86-10-64888708
20A Datun Road, Chaoyang District | Email: <a class="moz-txt-link-abbreviated" href="mailto:ccz@bao.ac.cn">ccz@bao.ac.cn</a>
Beijing 100101, China | WWW: <a class="moz-txt-link-abbreviated" href="http://www.lamost.org/~cb">www.lamost.org/~cb</a>
===============================================================</pre>
</body>
</html>