Skip to main content

ColonyMind

A mind for Hermes agents: persistent memory, commitments, initiative and measured autonomy.

Not just remember: keep promises and take the next step, within limits the owner sets. ColonyMind gives a personal Hermes agent memory across sessions and channels in people's own words, commitments it keeps, people it knows, and a small mind that acts on its own inside an autonomy level, budgets and quiet hours the owner sets. Asks reach the owner as plain messages, answered with a plain yes, a reply or a thumbs-up; the model can never approve anything.

This package is the ColonyMind sidecar and its colonymind command. It attaches to stock Hermes through the companion colonymind-hermes adapter, which colonymind init installs into Hermes' own environment; the sidecar runs in an environment of its own. No fork, no patches, no graph database, no extra key.

Install

You need Python 3.11 to 3.13 (3.12 recommended), a stock Hermes install (hermes-agent >=0.21.3,<0.24) whose hermes executable is on PATH, and one OpenAI-compatible chat endpoint, which Hermes already has.

uv tool install --python 3.12 colonymind
colonymind init
hermes gateway restart

pipx works too (pipx install --python python3.12 colonymind). ColonyMind becomes Hermes' memory provider: init names the provider it replaces, and colonymind init --uninstall puts it back. colonymind doctor checks the install, and colonymind demo shows the loop with no model, no keys and no network: a fact recalled across sessions, a promise followed up on its own, and an ask the person answers in plain words.

One container runs stock Hermes and ColonyMind together: ghcr.io/aevonix/colonymind, stock Hermes' own published image pinned by digest and patched nowhere (the container guide).

The vector store (LanceDB) is part of every install: semantic recall needs only an OpenAI-compatible embeddings endpoint (colonymind init --embed-url auto looks for one on the model endpoint). Where the endpoint has none, the vectors extra lets the sidecar run a pinned embedding model itself (uv tool install --force --python 3.12 'colonymind[vectors]', which brings PyTorch, then colonymind init --embed-url local, which downloads the model once into the instance directory).

To update: uv tool upgrade colonymind (or pipx upgrade colonymind), colonymind upgrade, hermes gateway restart.

Learn more

Licensed under the included MIT license. Optional dependencies retain their own licenses.

Metadata

Release files for colonymind 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 colonymind 1.0.0
File Size Uploaded Provenance Actions
colonymind-1.0.0.tar.gz 5.2 MB Without provenance Details for colonymind-1.0.0.tar.gz

Built distribution (wheel)

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

Total release size: 8.0 MB

Release files / colonymind-1.0.0.tar.gz

Download URL colonymind-1.0.0.tar.gz
Size 5.2 MB
Tags Source
SHA-256 checksum
How to use checksums
df2e65bbecbcc4fd935c34733181e286a7afdde7bf69df3b0a968e2728ec5f7c
BLAKE2b-256 checksum
How to use checksums
46323129786fa4852c95ee85b2257080948748436cde8adfc8037fa51ee882af
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.13.16

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

Download URL colonymind-1.0.0-py3-none-any.whl
Size 2.8 MB
Tags Python 3
SHA-256 checksum
How to use checksums
aabb5f71446c600f9857b9258e01ed021695fde6675e609d8fcdb832ad59a615
BLAKE2b-256 checksum
How to use checksums
eb7fb83335b14dbd7e9a1b7b7a8c037f3d9ca24cb3bb006e6d807302524f5377
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.13.16

Security for colonymind 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.

Release history Release notifications | RSS feed

This release

1.0.0 This release

2 release files No provenance

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