Skip to content
DevAgentRadarDevAgentRadar
Sign in
Proposals/Pin models and tune reasoning effort per task
proposaleveryoneP5High impactGoogle Antigravity

Pin models and tune reasoning effort per task

Antigravity 1.1.5 adds stable model slugs, /effort and --effort controls, a redesigned /model picker with effort gauges, and per-agent model frontmatter. Use them to pin reproducible model choices and trade latency for depth on the fly instead of accepting defaults.

Why this loop

This release makes model identity and reasoning depth first-class and session-stable: slugs work in /model and --model, /effort and --effort switch effort variants with a timeline gauge and status-line badge, and custom agents can declare model (flash, pro, or inherit) in frontmatter. Teams that standardize slugs and effort per task type get repeatable CLI launches and subagent behavior, fewer “wrong model” surprises, and intentional latency-vs-depth tradeoffs. Reliability fixes (MCP embedded resources, permission parsing with quoted metacharacters, background-task lifecycle) further support leaning on MCP tools and longer sessions once model/effort policy is explicit.

Proposed actions

  1. Adopt stable model slugs in scripts and docs via --model and the /model picker so every session pins the same base model.
  2. Use /effort or --effort to raise reasoning depth for design/debug work and lower it for routine edits; confirm the status-line effort badge before long runs.
  3. Set model in custom agent frontmatter (flash, pro, or inherit) so subagents run at the intended tier instead of silently inheriting.
  4. Re-test MCP servers after upgrade: authenticate from /mcp, and verify tools that return embedded resources now surface text and inline media.
  5. Revisit permission allowlists for commands with quoted shell metacharacters (e.g. --grep="a|b") so pipelines no longer trigger spurious prompts.

Action pack

Paste into your agent or query via MCP (`get_action_pack`) — free, no extra AI cost

Google Antigravity / agent task

# DevAgentRadar → Google Antigravity

Goal: turn this release signal into a safe pilot plan for my stack.

## Context
Assistant: Google Antigravity
Proposal: Pin models and tune reasoning effort per task
Summary: Antigravity 1.1.5 adds stable model slugs, /effort and --effort controls, a redesigned /model picker with effort gauges, and per-agent model frontmatter. Use them to pin reproducible model choices and trade latency for depth on the fly instead of accepting defaults.
Primary source: https://github.com/google-antigravity/antigravity-cli/releases/tag/1.1.5

## Why it matters
This release makes model identity and reasoning depth first-class and session-stable: slugs work in /model and --model, /effort and --effort switch effort variants with a timeline gauge and status-line badge, and custom agents can declare model (flash, pro, or inherit) in frontmatter. Teams that standardize slugs and effort per task type get repeatable CLI launches and subagent behavior, fewer “wrong model” surprises, and intentional latency-vs-depth tradeoffs. Reliability fixes (MCP embedded resources, permission parsing with quoted metacharacters, background-task lifecycle) further support leaning on MCP tools and longer sessions once model/effort policy is explicit.

## Suggested actions
1. Adopt stable model slugs in scripts and docs via --model and the /model picker so every session pins the same base model.
2. Use /effort or --effort to raise reasoning depth for design/debug work and lower it for routine edits; confirm the status-line effort badge before long runs.
3. Set model in custom agent frontmatter (flash, pro, or inherit) so subagents run at the intended tier instead of silently inheriting.
4. Re-test MCP servers after upgrade: authenticate from /mcp, and verify tools that return embedded resources now surface text and inline media.
5. Revisit permission allowlists for commands with quoted shell metacharacters (e.g. --grep="a|b") so pipelines no longer trigger spurious prompts.

## Config surfaces this release may change
- subagent definitions — check your repo before applying
- MCP servers — check your repo before applying
- permission rules — check your repo before applying


Please:
1. Map the change to concrete pilot steps
2. Flag security / permission implications
3. Keep the pilot reversible
agentmcpmodelsecuritycliRelease source ↗

Your loop

This browser · no sign-in · not shared as “you”

Community signal · 0 approved · 0 deferred · 0 ignored

Anonymous public tally (one vote per visitor; changing choice updates your vote). Not moderated product truth.

Originating release signal

Google Antigravity1.1.5Jul 21, 2026

1.1.5

Added a `/effort` command to view and change the current model's reasoning effort, with a left/right timeline-gauge picker and a direct `/effort <level>` form so you can trade latency for depth on the fly. · Added an `--effort` flag to select a model's reasoning-effort variant when launching the CLI. · Added stable, user-facing model slugs that appear in the `/model` picker and are accepted by `--model`, so you can pin a specific model reliably across sessions. · +10 more changes

Verified excerpt — the source's own words
- Added a `/effort` command to view and change the current model's reasoning effort, with a left/right timeline-gauge picker and a direct `/effort <level>` form so you can trade latency for depth on the fly.
- Added an `--effort` flag to select a model's reasoning-effort variant when launching the CLI.
- Added stable, user-facing model slugs that appear in the `/model` picker and are accepted by `--model`, so you can pin a specific model reliably across sessions.
- Added a `model` option to custom agent frontmatter so an agent runs at a chosen model tier (such as `flash` or `pro`) when invoked as a subagent, defaulting to `inherit` (the parent's model).
- Redesigned the `/model` picker to group models by their base model and choose reasoning effort from a timeline gauge navigable with Left and Right, and added an effort badge to the status line for models that expose multiple effort variants.
- Improved the `/settings` (`/config`) panel by making it a bounded, scrollable list so it renders correctly in short terminals instead of overflowing, and stopped it from flickering when opening and closing dropdowns.

Excerpt ends here — this release continues at the source ↗.

Primary source ↗