foundry-story-flow

Guides building, deploying, and verifying a Castia-based Foundry hosted agent from local to Teams.

Updated Sep 8, 2026
One-click install
npx skills add https://github.com/sethjuarez/castia --skill foundry-story-flow-sethjuarez
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: foundry-story-flow
Source: https://github.com/sethjuarez/castia/tree/main/plugins/foundry-agent-starter-kit/skills/foundry-story-flow
Command: npx skills add https://github.com/sethjuarez/castia --skill foundry-story-flow-sethjuarez

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It removes the guesswork from taking a Microsoft Foundry hosted agent through its full lifecycle: running locally, deploying to Foundry, and completing the Microsoft 365 Copilot/Teams publish-and-hire check, with approval gates before any live work. ## Core Features & Use Cases - Castia-based scaffolding: Creates a Python hosted agent with main.py, .env.example, azure.yaml, and baseline agent configs using the Castia SDK by default. - Three-stage verification flow: Walks through Local (uv run, /readiness, /responses smoke tests), Foundry (azd provision/deploy with recovery guidance), and Teams (publish, approve, hire) gates. - Approval boundaries: Never automates tenant permission changes, Teams installation, or admin approval; it guides the user through those steps instead. - Use Case: You have a Foundry project with a deployed model and want to ship a hosted agent to Teams. This Skill scaffolds the agent, verifies it locally, deploys it with azd, and walks you through the Teams publish-and-hire check. ## Quick Start Ask the assistant to build and verify the local to Foundry to Teams story flow for a new Castia-based hosted agent in this repo.

Frequently Asked Questions about foundry-story-flow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I deploy a Python hosted agent to Microsoft Foundry?▼

Run azd deploy <service> --no-prompt from the agent folder after setting the Foundry project endpoint and model deployment name in the azd environment. If infrastructure is missing, run azd provision --no-prompt first, then deploy again.

How do I test a Foundry hosted agent locally before deploying?▼

Copy .env.example to .env, fill in FOUNDRY_PROJECT_ENDPOINT and AZURE_AI_MODEL_DEPLOYMENT_NAME, then run uv sync and uv run python main.py. Verify with GET /readiness and POST /responses, including a streaming request.

What do I need before starting the Foundry agent story flow?▼

You need one Foundry project with a deployed model: the project endpoint URL and the model deployment name. These values go into a developer-local .env file first and are only added to the azd environment when moving to hosted deployment.

Why does my hosted Foundry agent fail with session_not_ready or 424 FailedDependency?▼

This usually means the hosted container crashed at startup, often from dependency drift between the local SDK source and the version pinned in deployment requirements. Read the hosted session logs before changing deployment settings.

Can this skill automate Teams installation or admin approval for my agent?▼

No. Tenant permission changes, Teams installation, organization publication, and admin approval are never automated. The skill guides you through the three manual gates: publish in Foundry, approve the request in Microsoft 365, and hire the agent in Teams.