Skip to main content

teenyfactories

Agent runtime for TeenyFactories. Postgres-backed pub/sub, multi-provider LLM calls, vector search.

📖 Documentation: teenyfactories.github.io/core

pip install teenyfactories
import teenyfactories as tf

@tf.on_state('documents', 'loaded').do
def analyse(item):
    tf.collection('documents').set(item['key'], state='analysed')

while True:
    tf.run_pending()
    tf.sleep(1)

Writing a row into a (collection, state) makes every subscriber to that pair pick it up: (collection, state) is a FIFO queue, and the handler consumes a row by transitioning its state (or deleting it). That's the framework.

Public API

Group Names
Pub/sub on_state, run_pending (the only pub/sub primitive — no message bus)
Data collection (.set, .add, .get, .get_all, .remove, .count, .exists, .first, .state, .where, .limit, .vector_search), embed
LLM llm (fluent builder — .ask / .run_agent_loop), call_llm (LEGACY)
MCP add_mcp_server, add_mcp_tool
Schedule on_schedule.every(N).<unit>.do(handler)
Secrets / files secrets, bucket_store (+ BucketStoreError family)
Logging log_debug, log_info, log_warn, log_error, log_persona
Time / IDs get_timestamp, get_timestamp_utc, generate_unique_id
Debug breakpoint
Lifecycle / config sleep, shutting_down, FACTORY_NAME, AGENT_NAME, AGENT_SLUG, AGENT_ID, __version__

Anything not in this list is not part of the supported surface — don't reach into submodules.

How it plugs in

A factory is a directory:

my_factory/
├── factory.yml
└── agents/
    ├── poller.py
    └── enricher.py

Each agents/*.py runs in its own container based on ghcr.io/teenyfactories/agent:dev — which has this library pre-installed. Single-factory mode runs each file as a docker compose service; multi-factory mode runs them under the orchestrator. Either way, agent code is just import teenyfactories as tf + handlers + main loop.

Full setup, compose templates, environment variable reference: see the core repo.

Versioning

Semantic versioning; current release 1.0.0 (pip install teenyfactories). The version is declared in pyproject.toml and teenyfactories/__version__.py (kept in lockstep) and published to PyPI by CI on push to main, idempotent per version.

License

MIT.

Metadata

Release files for teenyfactories 1.0.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for teenyfactories 1.0.0
File Size Uploaded Provenance Actions
teenyfactories-1.0.0.tar.gz 96.5 kB Without provenance Details for teenyfactories-1.0.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for teenyfactories 1.0.0
File Interpreter ABI Platform Actions
teenyfactories-1.0.0-py3-none-any.whl Python 3 none any Details for teenyfactories-1.0.0-py3-none-any.whl

Total release size: 199.8 kB

Release files / teenyfactories-1.0.0.tar.gz

Download URL teenyfactories-1.0.0.tar.gz
Size 96.5 kB
Tags Source
SHA-256 checksum
How to use checksums
c350a68a8048dffceb8b88c3e096f5057f8762de59241a3f91c124579225b4a0
BLAKE2b-256 checksum
How to use checksums
a181e96cb42d82d8747d70f8f89b95d448a1d8b41582dc20b2c9cb09e80e5eed
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.11.15

Release files / teenyfactories-1.0.0-py3-none-any.whl

Download URL teenyfactories-1.0.0-py3-none-any.whl
Size 103.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
1a3df1186305f0e500db03d63745719e293c34fc035fda23c21bb017ece03be7
BLAKE2b-256 checksum
How to use checksums
357087b052f45c4cbe2ed7d23e3274df2c3ace6188b64d60919424f52b7a510b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.11.15

Security for teenyfactories 1.0.0

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.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page