feat: Update LangChain runners to implement Runner protocol returning RunnerResult

[jsonbailey]

Summary

• `LangChainModelRunner` and `LangChainAgentRunner` now formally inherit from the `Runner` protocol class.
• `LangChainModelRunner.run()` implements the unified `Runner` protocol; returns `RunnerResult` with `content`, `metrics`, `raw`, and `parsed` fields. Structured output is supported via the `output_type` parameter.
• `LangChainAgentRunner.run()` updated to return `RunnerResult`; populates `tool_calls` in `LDAIMetrics` from observed tool call names in the message response.
• Legacy `invoke_model()` and `invoke_structured_model()` removed — the deprecated type definitions (`ModelRunner`, `AgentRunner`, `ModelResponse`, etc.) are kept in `server-ai` for provider CI compatibility until a cleanup PR removes them.

Stacking

Stacked on top of `jb/aic-2388/openai-runner-protocol` (PR #149).

Test plan

• `make test-langchain` — all LangChain provider tests pass
• `make test` — all tests across server-ai, langchain, and openai packages pass

🤖 Generated with Claude Code

---

Note

Medium Risk
This is an API-shape change for LangChain runners (moving to Runner.run() + RunnerResult and removing old response types), which could break downstream integrations and metric parsing if any callers still expect the legacy methods/fields.

Overview
Updates the LangChain integration to the unified Runner protocol: LangChainModelRunner and LangChainAgentRunner now expose a single run(input, output_type=...) that returns RunnerResult (content, raw, optional parsed) instead of the legacy invoke_model/invoke_structured_model responses.

Adds structured-output support to LangChainModelRunner.run() via output_type and normalizes inputs (string vs List[LDMessage]), and enhances agent runs to capture tool call names into LDAIMetrics.tool_calls.

Docs and tests are updated to use LDAIClient.create_model/ManagedModel.run() and the new result shapes; LangGraph agent-graph running no longer returns evaluation results on AgentGraphResult (they’re only flushed/tracked internally).

^{Reviewed by Cursor Bugbot for commit 1e61d42. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

There are 2 total unresolved issues (including 1 from previous review).

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit cbc377d. Configure here.}