This PR adds Active Support instrumentation to capture the following Active Agent events:

• Core events: prompt.active_agent, embed.active_agent (excluding .provider.active_agent events)
• Streaming events: stream_open.active_agent, stream_close.active_agent, stream_chunk.active_agent
• Tool and Processing events: tool_call.active_agent
• Infrastructure events: process.active_agent

The raw events look like this:

Open questions

• Do we want to subscribe to .provider.active_agent events for individual API calls in multi-turn conversions? If so, we may want to unsubscribe from prompt.active_agent and embed.active_agent events to avoid duplicates. I'm not sure if that would prevent us from showing the total duration for prompts, though.
• What charts/data would be useful for the majority of users to see on a default monitoring dashboard? I'm not sure if a full dashboard for Active Agent makes sense yet.
• Do we need special handling for exceptions? Honeybadger reports unhandled exceptions in jobs and web requests automatically, so additional error reporting my be unnecessary—unless we can provide additional context or catch exceptions that are rescued internally. (For example, Sidekiq provides its own API for error handling.) We'll rely on Honeybadger's built-in exception handling.
• Do we want to collect metrics? We can create charts from the the values in the instrumentation events, but rolling up metrics can sometimes make charting easier (and users can fine-tune the data they wish to send). We may add metrics in the future if users need them (e.g. if the current events are insufficient for monitoring things like duration, usage, etc.).

Known issues

1. The prompt.active_agent event includes the full prompt text, which may result in higher data ingest. It also includes it twice, once in the top-level messages field, and again in the parameters.messages field. We could remove one or both of these in the future, but it would be better if Active Agent resolves the duplicate fields. Do users want to be able to inspect the messages data when monitoring prompts?
2. The prompt.active_agent event also does not include the usage which Active Agent describes in the docs.

3. During streaming responses, Active Agent emits a lot of stream_chunk.active_agent events. We're currently including them, but could remove them in the future. Are they useful?

TODO

• Add some simple unit tests to verify the plugin installation/subscriptions
• Create a Honeybadger automatic dashboard template We'll do this later.

Reference

• Active Agent instrumentation docs
• Active Agent error handling docs
• Honeybadger event reporting docs
• Honeybadger metric reporting docs
• Honeybadger plugin guide