You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
explicit model picker - aligned to sesson model #618
When dealing with tented stuff - we need to be able to select certain models. right now I had to go back through and change a bunch of stuff afterward. Is there away to make it easier to do this before?
GHCP suggested - In VS Code mode, Squad subagents inherit the current session model. The repo config now says GPT-5.4 only, but if someone manually changes the VS Code model picker and ignores the instruction, the platform can still run that session on a different model.
Thanks for this, @cwiza! You've identified something we absolutely believe in.
Brady's response says it perfectly: users should be able to pin a model to an agent, and Squad should respect that — full stop. This is a first-class feature, not an edge case.
Here's the good news: the SDK support already exists. Our model resolution pipeline has this locked in:
We're building squad config model command right now so you can configure this conversationally from the CLI — no JSON editing needed. And yes, you can also just tell your coordinator "always use Opus for Flight" and it persists.
This is a win for enterprise workflows where model choices matter. Thanks for pushing on this. ��
Issue:#618 — explicit model picker - aligned to sesson model Assigned to: Flight (Lead) Reason: No specific domain match — assigned to Lead for further analysis
@copilot evaluation: No strong capability match — routed to squad member.
Activity
bradygaster commentedon Mar 26, 2026
Thanks for this, @cwiza! You've identified something we absolutely believe in.
Brady's response says it perfectly: users should be able to pin a model to an agent, and Squad should respect that — full stop. This is a first-class feature, not an edge case.
Here's the good news: the SDK support already exists. Our model resolution pipeline has this locked in:
This means once configured, an agent's model choice takes priority. Perfect for tented/restricted models — explicit user configuration always wins.
Here's what you get today:
Your
.squad/config.jsoncan pin models:{ "defaultModel": "claude-sonnet-4.6", "agentModelOverrides": { "flight": "claude-opus-4.6", "fido": "claude-haiku-4.5" } }Here's what's coming:
We're building
squad config modelcommand right now so you can configure this conversationally from the CLI — no JSON editing needed. And yes, you can also just tell your coordinator "always use Opus for Flight" and it persists.This is a win for enterprise workflows where model choices matter. Thanks for pushing on this. ��
feat(cli): add squad config model command for model pinning (#618)
github-actions commentedon Mar 26, 2026
📋 Assigned to EECOM (Core Dev)
Issue: #618 — explicit model picker - aligned to sesson model
EECOM will pick this up in the next Copilot session.
github-actions commentedon Mar 26, 2026
🏗️ Squad Triage — Flight (Lead)
Issue: #618 — explicit model picker - aligned to sesson model
Assigned to: Flight (Lead)
Reason: No specific domain match — assigned to Lead for further analysis
@copilot evaluation: No strong capability match — routed to squad member.
Team roster:
squad:flightsquad:proceduressquad:eecomsquad:fidosquad:paosquad:capcomsquad:controlsquad:surgeonsquad:boostersquad:gncsquad:networksquad:retrosquad:incosquad:guidosquad:telemetrysquad:voxsquad:dskysquad:simssquad:handbooksquad:ralphsquad:copilotfeat(cli): add squad config model command for model pinning (#618) (#628