Start with a concrete task and inspect its evidence. Package names and access modes follow the work you want to do.
Reproducible offline example
Use PyCharter’s public API with a versioned contract. Inspect a rejected record and reproduce a reviewed correction.
PyCharter 0.0.67 · public Python API · synthetic records · no account or MCP required
Run the price exampleIllustrative local mock
Inspect the PyActuator messaging mock, policy refusal, and optional audit record. Follow the separate local Python/MCP quickstart.
No Slack message is sent. Audit persistence requires a configured writer.
Explore the messaging examplePlanned reference workflow
The intended workflow combines versioned configuration, your database, feasible scheduling, durable recovery, and independent freshness monitoring.
The offline example does not extract FMP data, publish to a database, or run a schedule.
Read the intended workflowThe PyCharter plugin bundles focused instructions, reviewed examples, schema references, and access to maintained Optophi capabilities. The coding agent helps clarify the task, proposes files, runs supported checks, and explains evidence for your review.
Plugin 0.1.0 validates local JSON batches with PyCharter 0.0.67. Installation, discovery, actual checks and resumption were tested in Codex CLI 0.162.1. Broader pipeline operations and other agent hosts retain their own acceptance gates.
Install the PyCharter pilotThe price example runs today through a public Python API in a local environment. No always-on PyCharter process is needed.
Get the working exampleAn integration mode for exposing supported tools to an agent. A local MCP server needs to run while the client uses it; it does not schedule batch pipelines.
See the separate MCP guideOptophi API access. Create a scoped validation key through Optophi onboarding and run a guided first task over MCP or REST. Existing provisioned keys remain supported. Local examples require no account.
Review operating responsibilities