Proposal: allow webhooks to choose between agent and bash

[jslvtr]

A nice workflow we've been using is to start ill-defined work using agents; but as the work becomes better defined and understood, it's usually possible to create a script that does the work instead of having the agent do the work.

For example: initially we asked the agent to transfer an issue from Linear to multica together with its assignee, description, comments etc based on various criteria.

After some time when it became apparent that the criteria that we'd improved over time were now static, we moved the work to a script.

As such our Multica webhook from Linear just asks the agent to run the script.

Same thing with e.g. our Linear triage (when issue is created, Multica agent investigates and posts a report to Linear). This is also a script, and the Multica webhook just asks the agent to run it.

Similarly with our workflow for posting PR comments into the Multica issue as a comment for the agents to pick up. Used to be an agent workflow, now is a script.

Hence I think it would be nice for us to be able to simply trigger a bash script via the multica daemon in a target runtime host.