fix(agents-api): Fix chat endpoint behavior

[creatorrr]

• feat: Fix type checks
• fix(agents-api): Fix chat endpoint behavior

---

🚀	This description was created by Ellipsis for commit dbb1b32

Summary:

This PR fixes chat endpoint behavior, updates type checks, refactors client usage, changes the default model, and enhances tests in agents-api.

Key points:

• Fix chat endpoint behavior in agents-api and agents-api/routers/sessions/chat.py.
• Update type checks in agents-api/activities/types.py.
• Refactor client usage in agents-api/activities/task_steps/__init__.py.
• Change default model to gpt-4o.
• Enhance type definitions and tests.
• Remove demo.py, dialog_insights, relationship_summary, and salient_questions activities and workflows.
• Replace julep_client with litellm.
• Update ChatInput and CompetionUsage types.
• Modify functions to accept multiple owners in agents-api/routers/sessions/chat.py.
• Adjust chat function to handle recall parameter.
• Update pytype.toml and add tests for non-recall chat messages.

---

Generated with ❤️ by ellipsis.dev

[ellipsis-dev]

👍 Looks good to me! Incremental review on b4d96dc in 47 seconds

More details

• Looked at 549 lines of code in 8 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. agents-api/agents_api/routers/sessions/chat.py:118

• Draft comment:
  The handling of the recall parameter in get_messages seems incomplete. There's no integration of the fetched documents (doc_references) into the chat flow or response which might be necessary based on the PR description. Consider implementing or clarifying how these documents affect the chat outcomes.

    # Integrate doc_references into the chat flow or clarify their usage

• Reason this comment was not posted:
  Confidence of 20% on close inspection, compared to threshold of 50%.

Workflow ID: wflow_2pNqyYbeV0VPgFel

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

👍 Looks good to me! Incremental review on c5f3cec in 32 seconds

More details

• Looked at 549 lines of code in 8 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. agents-api/agents_api/routers/sessions/chat.py:124

• Draft comment:
  The PR description suggests changes to chat endpoint behavior, but the actual implementation contains placeholders like raise NotImplementedError("Truncation is not yet implemented"). If these features are not yet ready, it might be better to update the PR description to reflect the current state of the implementation to avoid confusion.
• Reason this comment was not posted:
  Confidence of 0% on close inspection, compared to threshold of 50%.

Workflow ID: wflow_rMaT5JPnqFBtdPQ7

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

👍 Looks good to me! Incremental review on 31e0728 in 44 seconds

More details

• Looked at 680 lines of code in 12 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. agents-api/agents_api/worker/__main__.py:11

• Draft comment:
  The PR description mentions the removal of demo.py, but it is not shown in the PR diff. Please ensure that all intended deletions are included in the PR.
• Reason this comment was not posted:
  Confidence of 0% on close inspection, compared to threshold of 50%.

Workflow ID: wflow_ci00HGJGgszbw1gE

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

👍 Looks good to me! Incremental review on dbb1b32 in 40 seconds

More details

• Looked at 1292 lines of code in 20 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. agents-api/agents_api/activities/embed_docs.py:18

• Draft comment:
  Consider removing the default value for include_title in the embed_docs function to enforce explicit behavior control by the caller.

    include_title: bool,

• Reason this comment was not posted:
  Confidence of 0% on close inspection, compared to threshold of 50%.

Workflow ID: wflow_8l3P9CiR4hOVzUfm

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.