fix(tasks): accept pi model configuration

[jonathanlab]

Problem

Cloud Pi runs did not accept a model or thinking level at start. The API required an ACP adapter. New Pi tasks then used the default model.

Changes

• Accept Pi model and thinking configuration without an ACP adapter.
• Save this configuration in the run state before the sandbox starts.
• Reject thinking levels that Pi does not support.

How did you test this code?

• Ran Ruff, formatting, and Python compilation.
• Added API tests for Pi model configuration and invalid Pi thinking levels.
• The Django API tests did not start because this environment does not have sqlx.

👉 Stay up-to-date with PostHog coding conventions for a smoother review.

Automatic notifications

• Publish to changelog?
• Alert Sales and Marketing teams?

Docs update

No documentation update is needed.

🤖 Agent context

Autonomy: Human-driven (agent-assisted)

I used Pi coding agent tools. I used /improving-drf-endpoints and /writing-tests. The task-run serializer sets initial Pi configuration. The client uses Pi RPC for live and deferred changes.

[trunk-io]

✨ Submitted to Merge by @jonathanlab. It will be added to the merge queue once all branch protection rules pass and there are no merge conflicts with the target branch. See more details here.

[greptile-apps]

_{Reviews (1): Last reviewed commit: "fix(tasks): accept pi model configuratio..." | Re-trigger Greptile}