I’d love to see JSON output for hive functionality via the CLI

I’d love to see JSON output for hive functionality via the CLI. There’s not currently predictable output harnessing, so for example if I want swarm to do some analysis or a task and then update a Github issue, rather than ask it to do that via MCP / CLI agentically (unpredictable, unreliable - I want a repeatable, guaranteed process) I can just get the JSON output from the CLI command and have code post that Github issue.

With hive-mind on the other hand, the CLI doesn’t give me a way to extract the output of a task, or hook into milestones along a task, to achieve that reliable Github issue based flow.

So I want to use the power of hive-mind but swarm is the CLI API that currently gives the output capability that can be programatically consumed and worked with.

But there’s still a lot I don’t know about the rapidly evolving feature set of Claude Flow, so perhaps I’m mistaken.

Currently I have an NPM package project I'll soon publish which harnesses Claude Code and Claude Flow in a controlled sequence via Github issues (inspired by how you respond to issues with the swarm here), but I'm not able to use hive-mind features because where with swarm, I can use `claude-flow swarm "Make these changes and then summarize what you did in markdown format." --output-format json --no-interactive` and capture that output to a GitHub issue. There's no way for me to currently leverage the `hive-mind` CLI non-interactively with output capture to achieve this same system.

Start by locating the hive-mind CLI command and compare its non-interactive and output handling with `claude-flow swarm ... --output-format json --no-interactive`. Verify how task results and milestone events could be captured programmatically; done means a repeatable JSON output path for the requested hive-mind workflow.

Written by the indexing model from the issue text.