Getting Started
Core Container Housing all SolutionsMPS Agentic Services
Overview
Container type: service | Port: 8080
solutions-agents is an HTTP API service. It runs as a Kubernetes Deployment
with an ingress, listens on port 8080, and serves interactive
API docs at /docs when running locally.
Prerequisites
- uv — manages Python, the venv, and all
dependencies:
curl -LsSf https://astral.sh/uv/install.sh | sh - Docker — for image builds
If this repo consumes private packages, configure your registry read token
first — see CONTRIBUTING.md in the repo root for the one-time setup.
Set up a clone
git clone https://github.com/null-capital/solutions-agents.gitcd solutions-agents
doppler run --project global --config prd -- ./run sync./run sync creates the venv, installs all dependencies (uv sync --all-extras), and installs the git hooks. Prefer the raw tool? uv sync
does the dependency half on its own. For private packages, the standard
doppler run command routes GEMFURY_READ_TOKEN to uv without persisting it.
Run locally
./run localwhich runs, with hot reload:
uv run uvicorn solutions_agents.main:app --reload --port 8080The API is at http://localhost:8080 and interactive docs at
http://localhost:8080/docs.
Run with Docker
./run build # docker builddocker run -p 8080:8080 --env-file .env solutions-agents:devPublished images live at ghcr.io/null-capital/solutions-agents —
every merge builds and pushes one (see the
Deployment guide).
Verify
./run check./run check runs exactly what CI runs (lint + format + types + test),
so green locally means green in CI.
Next steps
- Configuration — environment variables and per-environment overrides
- Deployment — how a push becomes a running workload
- API Reference — generated from the docstrings in
src/ - Development — the day-to-day contribution workflow