agy

name

agy

description

Bounded one-shot Gemini sub-calls through the Antigravity CLI (`agy`): multimodal judgment over screenshots, generated images, UI states, and diagrams, read-only adversarial review lanes that return typed falsifiable findings as a second-model perspective, Nano Banana image generation and edit iteration feeding html-studio and mermaid work, broad synthesis across competing designs, ambiguity reduction into crisp constraints, alternate approaches and counterexample hunts, and redacted log or dataset distillation. Use when a task gains from an external second model, a critique or red-team stage wants an independent Gemini lens, or the request names Gemini, Antigravity, Google Ultra, or Nano Banana. gpt-5.6 offload belongs to the codex skill; delegation across Claude's own surfaces belongs to agent-dispatch.

[AGY]

Antigravity is an external Gemini call admitted only where it adds capability beyond the local toolchain. Print-only mode runs the wrapper — one bounded prompt in, one JSON receipt out — but it executes tools and lands file writes without any permission flag, so safety is prompt discipline: a review or judgment prompt names no save path and directs inspect-only work, and a generation prompt binds every save path to a throwaway scratch root, never a repo path or $HOME. --dangerously-skip-permissions stays unused; it buys nothing print mode lacks. Antigravity output is advisory until local source, official docs, MCP output, or user intent confirms it.

[01]-[ROUTING]

[SCRIPTS]:

• [01]-RUNNER: wraps the CLI this skill invokes; carries prompt and models subcommands.

[02]-[CAPABILITY]

[INDEX]	[TRIGGER]	[WHY_GEMINI]
[01]	Screenshots, generated images, UI states, visual diffs, diagrams	Native multimodal judgment over --add-dir files
[02]	Critique or red-team stage wanting an independent second reviewer	Read-only findings from a different model lineage
[03]	Image generation, mockup and diagram iteration	Native generate_image rendering (Nano Banana)
[04]	Competing designs, ambiguous tradeoffs, long-context synthesis	Independent perspective outside the main context
[05]	Vague requirements needing constraints and edge cases	Ambiguity reduction as a separate pass
[06]	Counterexamples, risk inventories, blind spots on gathered facts	Adversarial second reading
[07]	Redacted log or dataset distillation in a scratch directory	Pattern finding without polluting main context
[08]	Explicit requests for Gemini, Antigravity, or Google Ultra	User-directed routing

[03]-[MODEL_POLICY]

Wrapper defaults pin the strongest Gemini reasoning model at its highest tier; that pin is the carrier of a standing operator policy, never a tunable preference. A weaker or faster tier is admitted only where a specific use case proves zero signal sacrifice, and the burden of that proof sits on the routing decision, never on the default. Agents never choose models in ordinary use; models lists the live catalog for capability accounting and skill maintenance, and AGY_MODEL re-pins only when the policy owner rules a proven exception.

[04]-[INVOCATION]

Run from this skill directory:

uv run scripts/agy.py models
uv run scripts/agy.py prompt "Compare these two approaches and return the top 3 tradeoffs." --timeout 5m
uv run scripts/agy.py prompt "Assess the screenshot at /abs/path/shot.png and return concrete UI defects." --add-dir /abs/path --timeout 10m

Both halves bind the multimodal contract: --add-dir grants the bounded directory, and the prompt names each file by ABSOLUTE path — a granted directory with a bare filename in the prompt resolves nothing. --add-dir grants only directories that answer the prompt, and the only writes a wrapper prompt ever requests are generation artifacts under a scratch root.

--log-file routes the CLI log to an explicit path; a sandboxed or scratch-rooted environment passes a writable scratch path there instead of relying on the default log location. AGY_BIN overrides the binary path (default agy), AGY_MODEL overrides the pinned model, AGY_LOG_FILE sets the default log path, and AGY_PRINT_TIMEOUT overrides the default 5m timeout.

Each prompt run returns one JSON receipt:

{"op":"prompt","output":"..."}
{"op":"prompt","fault":"auth_required","detail":"..."}

Faults are binary_not_found, auth_required, quota_exceeded, or process_error. auth_required resolves through interactive agy in a real TTY with Google OAuth as the account holding the Antigravity subscription.

[05]-[PROMPT_CONTRACT]

• State the task, its context, and the exact output shape in one self-contained prompt; Antigravity sees only what that prompt carries.
• Carry the constraints that bind the answer: audience, files already inspected, limits, and facts not to assume.
• Ask for ranked options, deltas, typed findings, or a direct answer; open-ended commentary is never the request.

[06]-[REVIEW_LANE]

This second-perspective lane runs bounded critique and red-team: freeze the evidence, invert the objective, demand falsifiable findings. It inverts the every-agent-writes law at exactly this boundary — the Gemini reviewer inspects and returns typed findings, and a Claude writer adjudicates each finding against source before applying the fix; the repair rail never crosses into the lane.

• [EVIDENCE]: Subject arrives frozen — exact files under --add-dir, invariants under test, executed check output, known constraints.
• [CONTAMINANT]: A conversation transcript and a producer's self-justification never cross into the lane.
• [OBJECTIVE]: Prompt directs disproof — violated invariants, omitted consumers, false assumptions, failure paths, visual-contract breaks.
• [BURDEN]: Work stays naive until it survives the attack, so a clean verdict is earned by an attack that finds nothing.
• [FINDINGS]: Each finding carries {severity, invariant, evidence, failure_path, minimal_fix}.
• [REJECTED]: Praise, style preference, and ungrounded speculation are shapes the prompt names rejected.
• [ADJUDICATION]: Claude re-derives each finding from disk, fixes true positives at the root, and refutes false ones with source, never by vote.

uv run scripts/agy.py prompt "ROLE: read-only adversarial reviewer. SUBJECT: /abs/path/target. INVARIANT UNDER TEST: <invariant>. AUTHORITY: inspect only; propose, never edit. OUTPUT: a JSON array of findings only, each {severity, invariant, evidence, failure_path, minimal_fix}; reject praise, style preferences, and ungrounded speculation. Return [] only if a real attack finds nothing." --add-dir /abs/path --timeout 5m

[07]-[GENERATION]

Image generation is the default agent's built-in generate_image tool, fired from inside an ordinary print-mode prompt — no CLI flag, subcommand, plugin, or models row invokes it. Every generation prompt carries three obligations: name the tool and forbid code-drawing (an unnamed request lets the agent fall back to scripting Pillow when the render fails), state the visual spec, and give an absolute save path under a scratch root the caller creates. Each render also persists as <image_name>_<epoch-ms>.jpg under ~/.gemini/antigravity-cli/brain/<conversation-id>/.

uv run scripts/agy.py prompt "Use your generate_image tool (never code-drawing) to generate: <visual spec>, aspect ratio 16:9. Save to /abs/scratch/name.png." --add-dir /abs/scratch --timeout 5m

Everything routes through prompt prose; the agent maps it onto the tool's parameters: Prompt, AspectRatio (1:1 default at 1024x1024, plus 2:3, 3:2, 3:4, 4:3, 9:16, 16:9 — 16:9 renders 1376x768), ImagePaths (up to 3 absolute paths to edit, combine, or reference — the edit modality: name the source image by absolute path and state the change), and ImageName. No tier selector exists at any surface — not in the tool schema, the CLI flags, or the model catalog; the backend serves Nano Banana Pro where the account has access and the original Nano Banana otherwise, and --model picks only the reasoning agent driving the tool.

Every render passes two gates before use. magick identify proves a real raster at the expected dimensions — the payload is JPEG regardless of the requested extension unless the agent converts, so the filename is never format proof — and a multimodal read-back through this wrapper describes the artifact against the visual contract. A file around 1KB is the failure signature: the backend returned no image (CORTEX_STEP_TYPE_GENERATE_IMAGE: no image generated in response in the --log-file) and the agent code-drew a placeholder; the never-code-drawing clause plus the size gate catches it, and one re-run resolves a transient backend miss.

Generated mockups and diagrams are inputs to the realizing skill, never deliverables: agy drafts the spec, renders, and read-back-critiques, then html-studio realizes the page and mermaid-diagramming realizes the fence. A closed loop is the strongest pattern — visual contract, render, read-back critique against the contract, targeted correction — never a single "make it pretty" round.

[08]-[REFUSAL]

• [SECRETS]: OAuth codes, tokens, credential files, and unredacted sensitive logs never enter a prompt.
• [AUTHORITY]: Facts owned by local source, official docs, MCPs, or repo commands come from those owners, never from Antigravity recall.
• [ROUTINE]: Edits, formatting, git operations, package upgrades, and checks the local toolchain owns stay local.
• [SCOPE]: prompt and models are the wrapper's only surface; agent selection, background tasks, and shell-login subcommands stay outside it.

Interactive agy in a real TTY owns those outside surfaces — ongoing conversations, workspace tool permissions, resume, plugin management, and sandboxed project work; agy --help is the flag and subcommand contract. gsd-* personas are interactive-runtime composition only: they hang under plain print mode and emit process narration rather than bounded answers under --mode plan, so a bounded one-shot review or judgment call always runs the default agent with a strong prompt, never a gsd persona. GSD's artifact pipeline — mapper, researcher, planner, checker, executor, verifier — composes inside a live Antigravity session where each role consumes only the prior role's artifact.