Research Engine SDK
marginalia-ai-sdk is the standalone, typed contract for building plugins for
Research Engine. It contains static manifest
models, boundary DTOs, scoped client protocols, decorators, chunking helpers, and portable
contract tests. It does not install or import the Research Engine core.
Install
python -m pip install "marginalia-ai-sdk>=0.6,<0.7"
A plugin declares a static plugin.yaml inside its top-level import package and advertises
that package through the research_engine.plugins entry-point group. Plugin acquisition and
removal belong to pip, uv, or pipx; the running engine never installs plugin dependencies.
[project.entry-points."research_engine.plugins"]
example = "example"
schema_version: 2
plugin_id: example
requires:
core_api: ">=0.6,<0.7"
python: ">=3.11"
permissions:
network: none
filesystem: plugin_data
provides:
mcp_tools:
- id: example.search
entry: example.tools:search
description: Search the example source.
input_schema:
type: object
properties:
query: {type: string}
required: [query]
Plugin code executes in the core process after explicit operator approval. Scoped clients are the supported API boundary, not a security sandbox; install and enable only trusted plugins.
API compatibility follows the core minor release through 0.x. A plugin supporting core
0.6.x should depend on marginalia-ai-sdk>=0.6,<0.7.
See the architecture, changelog, and issue tracker. Licensed under Apache-2.0.
Release files for marginalia-ai-sdk 0.6.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| marginalia_ai_sdk-0.6.0.tar.gz | 18.3 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| marginalia_ai_sdk-0.6.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 38.9 kB
Release files / marginalia_ai_sdk-0.6.0.tar.gz
| Download URL | marginalia_ai_sdk-0.6.0.tar.gz |
|---|---|
| Size | 18.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
254d4da1584101cd3cc95eabdb41776d6f8526524a4eed0caf4c160d2e8c3777
|
|
BLAKE2b-256 checksum How to use checksums |
a631242c8a280eeeaf781cfc8d347fc047aaded6e964423d147345922a2ed847
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 18, 2026.
Transparency logRelease files / marginalia_ai_sdk-0.6.0-py3-none-any.whl
| Download URL | marginalia_ai_sdk-0.6.0-py3-none-any.whl |
|---|---|
| Size | 20.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
4e0324d7b862dffc36599b7b58d75caf87cfe1fdefbc49777e3f8c82c5e0f5e7
|
|
BLAKE2b-256 checksum How to use checksums |
b8cc59121a2ed93d9246ee6077ee5054c78a34b50fb4c73ec5a3e7f0c6a16e2d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 18, 2026.
Transparency log