alfredagain
Domain-agnostic LLM orchestration engine built on LangGraph.
Alfred is a multi-agent assistant framework. You implement a DomainConfig to teach it your domain — entities, subdomains, prompts, database adapter — and Alfred handles the orchestration: routing, planning, CRUD execution, entity tracking, prompt assembly, and conversation memory.
Install
pip install alfredagain
Quick Start
from alfred.domain.base import DomainConfig, EntityDefinition, SubdomainDefinition
from alfred.domain import register_domain
class MyDomain(DomainConfig):
@property
def name(self) -> str:
return "mydomain"
@property
def entities(self) -> dict[str, EntityDefinition]:
# Define your entities...
...
@property
def subdomains(self) -> dict[str, SubdomainDefinition]:
# Define your subdomains...
...
# Implement remaining abstract methods...
# Register and run
register_domain(MyDomain())
from alfred.graph.workflow import run_alfred
response, conversation = await run_alfred(
user_message="Hello!",
user_id="user_1",
)
What Alfred Provides
- Pipeline orchestration — Understand → Think → Act → Reply, with LangGraph state management
- Entity tracking — SessionIdRegistry translates UUIDs to simple refs (recipe_1, item_2)
- CRUD execution — Schema-driven database operations with middleware hooks
- Prompt assembly — Domain-provided personas, examples, and context injection
- Conversation memory — Turn summarization, context windowing, session persistence
- Model routing — Complexity-based model selection (mini for simple, full for complex)
- Streaming — 11 typed events for real-time UI updates
Architecture
Core is fully domain-agnostic. It never imports any domain package. Domains implement DomainConfig (75 methods — 23 abstract, 52 with defaults) and call register_domain().
See docs/architecture/ for detailed documentation:
overview.md— Architecture index + pipeline diagramdomain-implementation-guide.md— How to build a new domaincore-public-api.md— Entry points, extension protocols
Development
git clone https://github.com/jv92admin/alfred-core.git
cd alfred-core
python -m venv .venv
.venv\Scripts\activate # Windows
pip install -e ".[dev]"
pytest tests/ -v
License
MIT
Metadata
Release files for alfredagain 2.8.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| alfredagain-2.8.0.tar.gz | 444.8 kB | Without provenance | Details for alfredagain-2.8.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| alfredagain-2.8.0-py3-none-any.whl | Python 3 | none | any | Details for alfredagain-2.8.0-py3-none-any.whl |
Total release size: 654.7 kB
Release files / alfredagain-2.8.0.tar.gz
| Download URL | alfredagain-2.8.0.tar.gz |
|---|---|
| Size | 444.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
382fe7d46cda99558580d630fa621c37c7424a8628b01d61862912a16706eb38
|
|
BLAKE2b-256 checksum How to use checksums |
6a182e6420d7cefb203cb5e6e211b4a42ace1f82153e249bed3a5a800117f855
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.9
|
Release files / alfredagain-2.8.0-py3-none-any.whl
| Download URL | alfredagain-2.8.0-py3-none-any.whl |
|---|---|
| Size | 209.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
1b669887d5ca3ace31253c22d5013496a50ef5148beea168c51bd7b8d2f295e6
|
|
BLAKE2b-256 checksum How to use checksums |
7a1c23a3e5fdde13b42e7bb7b67300483327426253fe3e4d06fd6bc32e098f2e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.9
|
Security for alfredagain 2.8.0
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.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A 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.
To learn more about protecting yourself from malicious actors, visit our package security documentation.