Skip to content

About

QuantAIService: Quant project admission policies and ops tooling for the shared personal-ai-service task runtime (formerly AIAuditBridge).

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

QuantAIService

An independently packaged, project-scoped task runtime for AI work.

English | 简体中文

Status

Experimental. Use at your own risk. This project is not financial advice and does not guarantee any trading result. It is an operations tool; it does not place orders and does not grant trading authority.

What this is

QuantAIService runs project-scoped AI tasks. API tasks and persistent assistant tasks share one contract, and each repository that uses the service owns its own runtime, schemas, prompts, and source validation. The package provides the task server (quant-ai-service), a client (ai-task), and an offline scenario SDK for ScenarioTaskSpec/v1 and ScenarioInstance/v1 validated with JSON Schema.

What this is not

It does not own strategy prompts, source validation, execution, or adoption; those stay in the business repositories that own them. It is not a shared core for the repository family. It does not place orders or authorize work by itself.

Quick start

Requires Python 3.10 or newer.

python -m pip install .
python -m unittest discover -s tests/task_service -v

Run the task server with your own configuration. The bundled config/quant.example.json is disabled by default and is a starting point only.

Configuration

Variable Required Default Purpose
AI_SERVICE_URL Yes none Base URL of the task service.
AI_SERVICE_TOKEN No none Bearer token used by clients.
AI_SERVICE_AUDIENCE No none Audience for the workload-identity token. Required by the client_from_env() helper; optional for the ai-task CLI when AI_SERVICE_TOKEN is set.

With the ai-task CLI, set AI_SERVICE_URL and then either AI_SERVICE_TOKEN or AI_SERVICE_AUDIENCE. The client_from_env() helper used by the scenario pilot CLI instead requires AI_SERVICE_URL and AI_SERVICE_AUDIENCE; AI_SERVICE_TOKEN stays optional, and when unset the helper requests a GitHub workload-identity token for that audience.

Server configuration lives in a JSON file based on config/quant.example.json. It defines the listening resource, the SQLite state path, accepted token issuers, and the provider backends. Replace every CONFIGURE_* placeholder with your own value.

Repository layout

src/        task server, client, and scenario schemas
config/     example service configuration
deploy/     deployment templates
ops/        monitoring helpers
docs/       design notes and contracts
tests/      unit and contract tests

Contributing

See CONTRIBUTING.md.

Security

See SECURITY.md. Do not open a public issue for a vulnerability. Keep credentials, private data, and external service tokens out of Git and logs.

License

MIT. See LICENSE. The task-service component carries its own MIT license in LICENSE.task-service.

About

QuantAIService: Quant project admission policies and ops tooling for the shared personal-ai-service task runtime (formerly AIAuditBridge).

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages