[Feature Request] Auto-update 'updated_at' timestamp in metadata.json on file modifications

[Chu4hel]

Is your feature request related to a problem? Please describe.
Currently, when I create a new track using /conductor:newTrack, the metadata.json is created correctly. However, when the AI subsequently modifies files within that track (implementing features or fixing bugs), the updated_at field in metadata.json remains stagnant unless I explicitly instruct the model to update it or run a specific command again.
This leads to stale metadata, making it difficult to sort tracks by actual recent activity.

Describe the solution you'd like
I propose expanding the toolset/system instructions to ensure updated_at is refreshed whenever files in a track are modified.

There are two potential approaches:

1. Implicit Update (Middleware/Hook): If the extension wraps file write operations, it could automatically detect a write to /tracks/{id}/* and update the corresponding metadata.json timestamp without requiring the LLM to "remember" to do it.
2. Explicit Tool/Function: Expose a lightweight function (e.g., touchTrack(id)) or modify existing file-writing tools to accept a flag that triggers a metadata update. The system prompt would then need to be updated to encourage the model to call this when iterating on a track.

Additional context
Without this, the "Conductor" loses track of which tasks are currently active versus which ones haven't been touched in a long time, forcing the user to manually manage JSON timestamps.