Agent Nebula Platform
agent-nebula-platform is the reusable middleware between developer-facing SecureAgent, the official A2A Python SDK, and Agent Nebula SDK security primitives. It provides protocol integration, configuration, security composition and runtime services so agent authors can focus on skills rather than repeatedly implementing server lifecycle, authentication and observability.
Why use the Platform?
Applications can select capabilities through plugins rather than adopting a monolithic agent server. The Platform composes A2A client/server operations, security requirements advertised by Agent Cards, OAuth clients and protected resources, skill invocation, A2UI, optional Registry onboarding, and operational services. Protocol handling remains with the official a2a-sdk; cryptographic and OAuth standards primitives remain in agent-nebula-sdk.
SecureAgent / host application
|
agent-nebula-platform
|-- a2a/ A2A SDK integration and Agent Card security
|-- security/oauth/ OAuth client, registration and protected resources
|-- skills/ Transport-neutral skill dispatch
|-- a2ui/ Declarative UI and action integration
|-- integrations/ Optional Agent Nebula onboarding
`-- platform/ Lifecycle, plugins, configuration, telemetry, metrics
|
official a2a-sdk + agent-nebula-sdk
See Platform architecture and documentation index.
Protocols
A2A 1.0 is implemented using the official Python SDK, including client/server integration and security-scheme handling. A2UI 1.0 provides declarative UI data and actions over A2A; it is not a separate agent transport. MCP is planned, not yet implemented. See A2A, A2UI, and MCP status.
Security and authorization
The Platform interprets A2A Agent Card security declarations and composes supported authentication adapters with OAuth client and protected-resource services. It supports security integration for API keys, HTTP authentication, OAuth/OIDC and TLS/mTLS as configured by the host and advertised by the agent. Reusable OAuth, JOSE/JWT, JWK and DPoP primitives are delegated to agent-nebula-sdk; protected-resource authorization enforces exact required skill scopes. Optional Agent Nebula integration enables hosted registration, discovery, OAuth and policy-managed access; those hosted services are distinct from the local Platform package. See security, A2A security integration and Registry integration.
Plugins and lifecycle
Capabilities are assembled through RuntimePlugin and PluginRegistry, with pre-server and post-server initialization, required/optional failure handling, health/readiness, and reverse-order shutdown. Available plugin areas include skills, A2UI, OAuth clients, protected resources, client registration, Registry onboarding, metrics, telemetry and scheduling. See plugins and lifecycle for contracts and operational behavior.
Runtime configuration and filesystem permissions
The runtime path resolver provides configured-root containment, not an operating-system sandbox. runtime_path() confines paths to its configured runtime root; shared_path() may also permit the configured product root. The resolver does not enforce a global /run/agent-nebula-only rule, and Python code cannot prevent arbitrary filesystem access by other process components. For a deployment requiring writable runtime state only under /run/agent-nebula, configure that directory as the writable runtime mount and enforce restrictions with OS/container permissions and read-only mounts. Other paths may still be required read-only for Python packages, certificates and configuration. See runtime and isolation.
Observability and operations
The Platform provides plugin health/readiness, lifecycle events, metrics, telemetry, retry, scheduling and execution tracing. Enable and configure these capabilities in the host composition rather than assuming all are active by default. See runtime services and observability.
Install
Python 3.12 is required. The Platform supports Agent Nebula SDK 5.x. Install the latest published release:
pip install agent-nebula-platform
For examples and application-level usage, see the SecureAgent examples repository and the documentation index.
License
Apache License 2.0. See LICENSE.
Metadata
Release files for agent-nebula-platform 5.0.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| agent_nebula_platform-5.0.1.tar.gz | 182.5 kB | With provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| agent_nebula_platform-5.0.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 377.7 kB
Release files / agent_nebula_platform-5.0.1.tar.gz
| Download URL | agent_nebula_platform-5.0.1.tar.gz |
|---|---|
| Size | 182.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
2d28ed6398955b5367fb39aefa782740509d84923f4387d99b51bdb091b6369d
|
|
BLAKE2b-256 checksum How to use checksums |
d6728cdeb047e3ff9b67a09637fa0707ffb020933ee4667aff1cfae8dd401a12
|
| 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 Oct 19, 2026.
Transparency logRelease files / agent_nebula_platform-5.0.1-py3-none-any.whl
| Download URL | agent_nebula_platform-5.0.1-py3-none-any.whl |
|---|---|
| Size | 195.2 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
9d6f8f21a5111e8b8ff88c8ba8471e5e5130df5aa2adddc3ac4fb1f4a6d006e3
|
|
BLAKE2b-256 checksum How to use checksums |
f0ce6daa1ebb3c8ff4e805934dd268c9bd49b54ab87a548cbfb2265112007964
|
| 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 Oct 19, 2026.
Transparency logSecurity for agent-nebula-platform 5.0.1
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.
Full provenance
All 2 files in this release have been uploaded with provenance metadata.
Attestation for agent_nebula_platform-5.0.1.tar.gz
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 Oct 19, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.