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: Hide model-only skills with disable-slash-command and script print-mode slash output
Summary: Antigravity 1.1.12 lets teams declutter `/` via SKILL.md `disable-slash-command: true`, script read-only slash commands with `-p` (TSV/JSON, no agent turn), and capture clean `models`/`agents` lists via `--output-format json`.
Primary source: https://github.com/google-antigravity/antigravity-cli/releases/tag/1.1.12

## Why it matters
Large skill libraries were flooding the slash menu even when users never type those skills by name. Frontmatter `disable-slash-command: true` keeps a skill model-discoverable and invocable while removing it from `/` and `/name` resolution—ideal for bulk/internal skills. Print mode now answers `/permissions`, `/hooks`, `/help`, `/changelog`, and `/config` non-interactively (one TSV record per line, or structured under `--output-format json`/`stream-json`) without starting a turn, spending quota, or leaving a conversation; `/help` lists exactly which commands print mode answers. `models` and `agents` gained the same `--output-format` flag with spinner/errors off stdout so captures stay pure lists. Pair that with the artifact viewer outline (`t`) when reviewing long release or design markdown. Adopt these in shared skill repos and automation before the menu and ad-hoc CLI scraping become bottlenecks again.

## Suggested actions
1. In every SKILL.md that should be model-invoked only (not typed via `/`), add frontmatter flag `disable-slash-command: true` so the skill stays discoverable by the model but is hidden from the `/` menu and `/name` resolution.
2. Run print mode for inventory without an agent turn: `-p "/help"` (tab-separated lines); repeat for `-p "/permissions"`, `-p "/hooks"`, `-p "/config"`, and `-p "/changelog"`, or add `--output-format json` / `stream-json` for structured payloads.
3. Capture clean catalogs for scripts: `models --output-format json` and `agents --output-format json` (stdout is only the list; errors and the progress spinner no longer pollute captures).
4. When reviewing long markdown artifacts (changelogs, designs), open the artifact viewer heading outline with `t` and jump to sections instead of scrolling.
5. Re-enable heavy MCP/plugin/skill setups that previously failed tool-declaration limits—the per-session tool declaration ceiling was raised in 1.1.12.

## Config surfaces this release may change
- background and headless runs **(high confidence)** — check your repo before applying
- skills — check your repo before applying


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