dp-engineering-assistant
Allows discovery and inspection of CE-RISE repositories hosted on Codeberg, and checks for updates from curated documentation and release channels.
Provides deployment artifact generation for Docker Compose, producing starter files derived from the CE-RISE GitOps template.
Enables checking release/tag channels on GitHub, discovering model repositories under the configured namespace, and integrating with GitHub Container Registry for OCI images.
Provides deployment artifact generation for Kubernetes manifests, producing starter files derived from the CE-RISE GitOps template.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@dp-engineering-assistantassess our readiness for digital passport adoption"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
CE-RISE Digital Passport Engineering Assistant
A repo-native stdio MCP server that helps AI agents and human users make better use of existing CE-RISE Digital Passport components, services, schemas, documentation, and deployment workflows.
This assistant is not a replacement for CE-RISE assets. Its role is to guide discovery, configuration, implementation planning, readiness checks, and appropriate reuse of the existing CE-RISE ecosystem.
The main workflow is general Digital Passport adoption: helping an organization understand compliance drivers, value-chain information flows, implementation readiness, and value opportunities from shared information, regardless of company size or organizational label.
The main human-facing CE-RISE solution entry point is:
The public documentation for this MCP server is published at:
What this repository contains
A minimal stdio MCP server in
server/mcp_server.pyA deterministic initial CE-RISE capability and component catalog in
data/solution_catalog.jsonA curated connected-source manifest in
data/connected_sources.jsonA read-only live service connection manifest in
data/live_service_connections.jsonA deployment artifact manifest in
data/deployment_artifacts.jsonA reference-example generalization manifest in
data/reference_examples.jsonAn update channel manifest in
data/update_channels.jsonMCP registry metadata in
server.jsonA container runtime skeleton in
DockerfileLocal run, smoke, and validation scripts in
scripts/A first deterministic design workflow example in
examples/01-design-workflow/Documentation in
docsusing BookGitHub mirror automation for release, OCI image publication, and MCP registry publication
Codeberg Pages automation for the documentation site
The server is intentionally repo-native at this stage. It does not introduce an installable Python package.
Related MCP server: bunge-ds-mcp
MCP Tool Shape
The initial tool surface is catalog-backed and deterministic:
list_solution_capabilitieslist_solution_componentsmap_user_goal_to_ce_rise_capabilitiesrecommend_passport_architecturegenerate_implementation_planassess_implementation_readinessassess_adoption_contextmap_value_chain_flowsidentify_value_opportunitiesrecommend_adoption_pathgenerate_implementation_roadmaplist_deployment_artifact_templatesgenerate_deployment_artifact_plangenerate_deployment_artifactsassess_deployment_artifact_readinesslist_reference_examplesgeneralize_reference_examplelist_update_channelscheck_update_channelsbuild_update_aware_solution_contextdiscover_model_repositorieslist_connected_sourcescheck_connected_sourcesinspect_connected_sourcebuild_connected_solution_snapshotlist_live_service_connectionsprobe_live_serviceinspect_live_service_openapibuild_live_service_readiness_snapshot
The initial resource surface is:
ce-rise://solution/catalogce-rise://solution/scopece-rise://sources/manifestce-rise://services/live-connectionsce-rise://deployment/artifact-templatesce-rise://examples/reference-generalizationce-rise://updates/channels
The first implementation is documentation/resource-oriented with a small read-only service discovery layer, a deployment-artifact output layer, a reference-example generalization layer, and an optional update-awareness layer. It connects to curated CE-RISE repositories and documentation references, can inspect local sibling repositories when they are available, can optionally check remote repository/documentation URL availability, can probe read-only live service endpoints such as health, readiness, version, and OpenAPI, can return Docker Compose or Kubernetes starter file contents derived from the existing CE-RISE GitOps template, can generalize reusable patterns from the local demonstrator without copying its concrete demo data or local-only settings, and can check configured release/tag/documentation channels for current upstream metadata when requested.
Normal component-aware outputs include an update_awareness field. This makes update needs visible even when the user did not explicitly ask about versions: the field lists relevant configured update channels and recommends the update-aware tools before implementation, deployment, or detailed technical decisions.
Generation tools can carry update context directly. For example, generate_deployment_artifacts accepts check_remote_updates; when enabled, it checks configured update channels, includes version_context, writes a generated VERSION-CONTEXT.md file into the returned artifact set, and uses checked model_artifact_version tags where they map safely to generated registry catalog entries.
The update layer covers software components, reference examples, deployment templates, documentation availability, and known data model repositories through explicit read-only channels. The server can also discover additional CE-RISE model repositories under the configured model namespace with discover_model_repositories; newly discovered repositories are returned as candidate update channels for review instead of being silently trusted as implementation inputs. Template and model-architecture repositories can be tracked for version awareness without being treated as deployable data model artifacts.
It does not call live CE-RISE business operations such as record creation, validation, query, indicator computation, or registry refresh.
Adoption, architecture, value-chain, and roadmap outputs include source_references and live_service_connections fields where relevant, so recommendations are grounded in CE-RISE assets rather than presented as standalone advice.
Deployment artifact outputs are starter planning artifacts only. They do not replace the canonical dp-system-gitops-template repository or the component-specific deployment documentation.
Compliance-oriented outputs are planning aids only. They do not provide legal certification or legal advice.
Local Use
Run the MCP server:
./scripts/run-local.shRun a local MCP smoke check:
./scripts/smoke-mcp.shRun a container MCP smoke check:
./scripts/smoke-container.shRun local validation:
./scripts/validate-local.shEquivalent make targets are available:
make smoke
make smoke-container
make test
make validateMinimal local MCP client configuration:
{
"mcpServers": {
"dp-engineering-assistant": {
"command": "/home/riccardo/code/CE-RISE-software/dp-engineering-assistant/scripts/run-local.sh"
}
}
}Publication Model
Development happens on Codeberg:
The documentation site is published from the Codeberg repository:
The GitHub repository is a mirror used for release-side automation, following the CE-RISE MCP publication pattern:
build and publish the MCP OCI image to GitHub Container Registry;
update
server.jsonfor the release version and image tag;publish the MCP server metadata to the official MCP Registry using GitHub OIDC;
create GitHub releases for Zenodo integration.
After release, the server is expected to be reachable in two ways:
MCP Registry and OCI package: MCP clients can discover the stdio server by the registry identity below and run the published GHCR image.
MCPO/OpenAPI gateway: when registered behind the CE-RISE MCPO gateway, HTTP/OpenAPI clients can reach the same MCP tools through the gateway endpoint published in the project documentation. The gateway URL is deployment metadata and is not hardcoded in this repository.
The MCP registry identity is:
io.github.CE-RISE-software/dp-engineering-assistant
The OCI image pattern is:
ghcr.io/ce-rise-software/dp-engineering-assistant-mcp:<release-version>
License
Licensed under the European Union Public Licence v1.2 (EUPL-1.2).
Contributing
This repository is maintained on Codeberg as the canonical source of truth. The GitHub repository is a read mirror used for release archival, OCI image publication, MCP registry publication, and Zenodo integration. Issues and pull requests should be opened on Codeberg.
Funded by the European Union under Grant Agreement No. 101092281 - CE-RISE.
Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the granting authority (HADEA). Neither the European Union nor the granting authority can be held responsible for them.
(c) 2026 CE-RISE consortium.
Licensed under the European Union Public Licence v1.2 (EUPL-1.2).
Attribution: CE-RISE project (Grant Agreement No. 101092281) and the individual authors/partners as indicated.
Developed by NILU (Riccardo Boero - ribo@nilu.no) within the CE-RISE project.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/CE-RISE-software/dp-engineering-assistant'
If you have feedback or need assistance with the MCP directory API, please join our Discord server