mORMot Open Source

Why your function returns False
•  Assumes signature is always raw 64 bytes (R||S)
•  Your check Length(...) <> SizeOf(TEccSignature) exits early for any DER signature. OpenSSL (and many toolchains) output DER. So you never reach Verify for DER input.
•  Public key format mismatch
•  lCertificate.GetPublicKey returns the X.509 SubjectPublicKey (SPKI) in ASN.1. TCryptAsym.Verify expects EC public key in one of a few raw/PEM/DER encodings it can parse; SPKI isn’t always accepted directly by its low-level PemDerRawToEcc(). You either need to rewrap the SPKI to the EC compressed key DER, or use ICryptCert.Verify (which does that internally).
•  Data bytes/encoding

---

read code in gist

credit: gpt

ps: I tested the code

@ab: do you think the ai-model hallucinated?

You are missing the object creation that will trigger an AV and use ObjectToJson, not JsonEncode.

function TUtilisateursService.GetUtilisateur(aID: TID): RawJSON;
var
  obj: TOrmUtilisateurs;
begin
  obj := TOrmUtilisateurs; // <- missing create
  try
    if not fClientDB.Orm.Retrieve(aID, obj) then 
      raise ERestException.Create('404 - Utilisateur non trouvé');

    Result := ObjectToJson(obj); 
  finally
    obj.Free;
  end;
end;

ou

function TUtilisateursService.GetUtilisateur(aID: TID): RawJSON;
var
  obj: TOrmUtilisateurs;
begin
  obj := TOrmUtilisateurs.Create(fClientDB.Orm, aID);
  try
    if ID = 0 then
      raise ERestException.Create('404 - Utilisateur non trouvé'); 

    Result := ObjectToJson(obj); 
  finally
    obj.Free;
  end;
end;

read there: SAD1.18 5.2. Working with Objects

If it helps you find the way, on my side, I use TWebSocketProtocol.UpgradeBearerToken for that: read more there

don't be fooled ..

.. and it's not the only data source; You know.. personal privacy.. checkbox.. big-data, etc...

The main issue, IMO, comes from when most datasets were built. The first public release of mormot v2 was around late 2020 (first commit in march 2020), but by then the models had already ingested mormot v1 source code. That’s why on gpt-3, gpt-4 (and even gpt-5!), you often get v1-based code in the answers - I'am quite sure you got the TSqlModel instead of TOrmModel things, etc.  which is why the “big prompt” is actually useful.

Also, if you let your tools adapt to your own habits, you’ll naturally get better results. On my side, no matter what I ask, I always end up with a marmot somewhere in the answer

No. At this moment, the mcp-delphi server only allows AI agents to use your IDE to compile your project (Delphi, Lazarus, or FPC directly) and get the output, nothing more. That said, if your coding tool supports "auto-mode", you can let the agents create, compile, and fix the project for you - while you’re cooking, for example I tried it with gpt-5, project built successfully and potatoes where not burnt lol.

Then I thought it could be useful to have a dedicated MCP server for working with mormot. It would definitely reduce the number of tokens consumed, since the agents would be able to query a well structured knowledge base for the relevant part of the framework instead of injecting a huge prompt. I also believe the agents would "think" better as their memory context wouldn’t be overloaded.

Thanks! I’ll review it this weekend. Is it based on the latest mORMot update?

I published a MCP server for delphi and lazarus/fpc; I am working on a special setup focused on mormot2. I must say from my tests with gemini - and now gpt-5 included - it's quite amazing. Token usage will be reduced a lot compared to "the-big-prompt".

You can find it there: https://github.com/flydev-fr/mcp-delphi

at least I get some more doc to ingest.

Aw nasty. Thanks you ab

I just tested with `curl http://localhost:9088/api/main/UserService.ReadTest?user_uuid=3C28EA56-37FE-4ABC-A0B7-508D8C634E21` and I have the very same issue, result:

uuid is: 17868178-0000-0000-0000-000000000000

Then I have also compiled the server on ubuntu and same problem. FPC version on Ubuntu is 3.3.1. The lazarus project file is based on MVCServer.lpi with -Xm and compiled without optimization. I will test it with a simple and new project..

---

Edit: 

I have created a new simple project on ubuntu and compiled, result is the same, the guid is wrong.

Again, just a small question about TGUID as I’m hitting a cross-platform snag with FPC 3.2.2 but with TGUID defined as param on SOA:

IUserService = interface(IInvokable)
['...'] 
  function Read(const user_uuid: TGUID): TServiceCustomAnswer;
end;

- Win64 build (FPC 3.2.2): user_uuid arrives with the expected value.
- Linux x86-64 build (same sources & FPC version): user_uuid is always null.

Is this due to missing record RTTI for TGUID under FPC on Linux?
I can tackle the problem by passing it as RawUtf8, but I'am asking for my own knowledge.

thanks

@Alek, to force partial changes to be reloaded, apply theses modifications:

on mormot.rest.mvc.pas#L181 add:

FViewPath: RawUtf8;

and the public property:

property ViewPath: RawUtf8 read FViewPath write FViewPath;

on line mormot.rest.mvc.pas#L1245 add:

{$ifdef DEBUG}
fViewPartials.ViewPath := folder;
{$endif}

replace mormot.core.mustache.pas#L1317-L1332 with:

function TSynMustachePartials.GetPartial(
  const PartialName: RawUTF8): TSynMustache;
var
  i: PtrInt;
  Temp: TSynMustache;
begin
  if self=nil then begin
    result := nil;
    exit;
  end;
  i := fList.IndexOf(PartialName);
  if i<0 then
    result := nil
  else 
  begin
    // Partial files by default in mORMot are loaded just once and then cached which is very bad
    // during debugging when files are changed and it requires constant restart of the application.
    // With this change we're disabling cache in debugMode: -dDEBUG on fpc
    Result := TSynMustache(fList.Objects[i]);
    {$ifdef debug}
    begin
      Temp := TSynMustache.Parse(AnyTextFileToRawUTF8(UTF8ToString(MakePath([FViewPath, PartialName + '.partial'])), true));
      if Temp.Template <> '' then
        Result := Temp;
    end;
    {$endif debug}
  end;

@ab, any chance you consider adding it or similar - or did we missed something?

---
Edit: I missed last post of @ab - watching for files changes and triggering FlushAnyCache() should be the right way to do it while developping. Anyway, the method presented here still make getting partial markup changes easier.
---

I made a gist trying to explain it (should be correct), read there: https://gist.github.com/flydev-fr/e0f0a … c2ac78f2bc

Call FlushAnyCache when you change data in your model. If I am correct, on the blog sample, the default data is cleared and filled with new data again if empty on GetViewInfo(). The method is virtual so you can implement your own flush logic there.

I can't reproduce it - it work as expected at least on fpc 3.2.2.

edit: ... and mormot 2.3.10775

TRestClientDB use an hidden TRestServerDB; you could map to an external db for example this should work:

Props := TSQLDBZEOSConnectionProperties.Create(...);
Props.ThreadSafeConnection.Connect;
OrmMapExternal(aORMSQLModel, [TYourOrmMapped], Props);
ORMDB:= TRestClientDB.Create(aORMSQLModel, nil, ':memory:', TRestServerDB, False, StringToUtf8(aPassword));
TRestServerDB(ORMDB.Server).Server.CreateMissingTables(0, [itoNoAutoCreateUsers]);

My pleasure

I forgot to talk about the tool. To get the mopenapi tool, just compile it from there src/tools/mopenapi (blog post) and then you will be able to get a JsonClient and DTOs defintions in a second and ready to work with your API, then just throw the DTOs code on Gemini if you want them on your interfaces or whatever.

To get the JSON file from YAML, just use any YAML <> JSON converter - I will send a PR to handle YAML on mopenapi tool directly if ab accept it

Hi @zen, yes I can - let me create a repo for that.

Yes because of two things. The first is that the frameworks is complex and I know models are subject to hallucinations while working with mormot; For example you can get GPT working on v2 only after corrections and sending to the models samples.

The second, I had the sentiment that Gemini had a lot of power, and with 1 million of tokens context (like GPT 4.1 now) I tested to throw the full source code. And it worked like a charm

And as the tests are written by @ab and show the right way of implementing things, then I through it was the best idea to pack the tests. FYI, packing the WHOLE framework don't work as you will get too much tokens, like 2 or 3 millions - so you must choose the best parts based on what you are trying to achieve.

Lol 

yes! I did a try last month with Gemini, to generate a full project based on mormot2. I packed almost the whole mormot source and test folder with repomix, added some custom corrections I was used to and thrown a prompt of 900,000 tokens to Gemini. It has ended with a project of multiple REST servers, all the interfaces and services implementations and an openapi definition file.
I then could generate the whole JsonClient and DTOs with mormot tool. I just had to make some minors corrections - impressed as f**.

I forgot to tell that if you want to play with it, I ported the project on mORMot v2 some months ago and I can publish it.

I published a small example on github: flydev-fr/mormot2-jsonclient
and small comments project1.lpr#L29-L38

look how the client is used in `petstore.client` (generated by mopenapi)

Check request headers:

I am a bit late, for question #1 I like to give as reference this awesome post:

refresh-tokens-what-are-they-and-when-to-use-them on auth0.com blog.

Well explained and make things crystal clear.

Also, it will be easier to find it when searching on forum

hi, hard to say without more details.  Start by enabling log with LOG_VERBOSE level then check logs once it happen again.

@youssef, remove route and add the correct one with https scheme. Your issue seem to be the route is already registered for http.

read this:

link to documentation + windows tool:   
https://synopse.info/forum/viewtopic.ph … 368#p42368

I think @zen010101 is suggesting you this post:
https://synopse.info/forum/viewtopic.ph … 664#p27664

@cadnan, see gpt o1 suggestions:

https://gist.github.com/flydev-fr/844bd … a09a01153f