rescue

A command for handing a task to the Antigravity command-line tool, with options for background jobs, deadlines, resuming conversations, and safety checks.

In plain words
What is it for?
Use it to run a task now or in the background, wait for completion, resume a previous Antigravity conversation, or start a fresh one.
Why use it?
It lets you start long or risky work without losing track of it, while keeping a job record and allowing you to inspect or stop the work later.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/limeflash/antigravity-plugin-cc/rescue
Clone the repo
git clone --depth 1 https://github.com/limeflash/antigravity-plugin-cc
Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,313 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00035 $0.01313
Opus 5 $0.00017 $0.00656
Sonnet 5 $0.00007 $0.00263
Haiku 4.5 $0.00003 $0.00131

Measured yesterday against content hash 0ab7edf3fc37, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

rescue scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/agy/commands/rescue.md · 117 lines

How it starts

The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Hand the user's task to the Antigravity CLI (agy) through the plugin's Node.js companion, which adds background execution, job control, and persistent state under <repo>/.agy-plugin/.

Use /agy:rescue instead of /agy:delegate when the user wants to:

  • Start the work and come back later (--background).
  • Block on a long-running task with a deadline (--wait; default 10 minutes timeout, configurable in agy-companion.mjs).
  • Resume the most recent agy thread (--resume).
  • Force a brand-new conversation regardless of history (--fresh).
  • Get a job id they can later inspect with /agy:status, fetch with /agy:result, or stop with /agy:cancel.

For one-shot, "ask agy a question and read the reply" UX, prefer /agy:ask (synchronous, Bash wrapper, no Node required).

How to invoke

The user's request (treat as opaque text — pass it as a single shell-safe argument; do not interpolate or splice it into the command, do not strip characters like ", $, ;, \ or backticks):

$ARGUMENTS

Lift any leading routing flags out of $ARGUMENTS and place them before the task text on the companion command line:

  • --background, --wait, --resume, --fresh: boolean
  • --model <alias> or --model=<alias>: value

Use the Bash tool to run:

node "${CLAUDE_PLUGIN_ROOT}/scripts/agy-companion.mjs" rescue [flags...] "<task-text>"

…with <task-text> quoted as one shell argument so characters like ", $, ;, \ and backticks cannot break out.

Examples (after you've parsed the flags out):

node "${CLAUDE_PLUGIN_ROOT}/scripts/agy-companion.mjs" rescue "investigate the flaky test in src/db/queries.test.ts"
node "${CLAUDE_PLUGIN_ROOT}/scripts/agy-companion.mjs" rescue --background "audit the auth layer for missing input validation"
node "${CLAUDE_PLUGIN_ROOT}/scripts/agy-companion.mjs" rescue --background --wait "fix the off-by-one in pagination"
node "${CLAUDE_PLUGIN_ROOT}/scripts/agy-companion.mjs" rescue --model opus "rewrite the cache layer to use Redis"

Model aliases match the Bash wrapper: flash-low, flash-medium, flash, pro-low, pro, sonnet, opus, gpt-oss. The companion passes the alias through to the wrapper's per-call patch of settings.json; no change to the TUI's selected model.

Read the full file on GitHub · 117 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 117 lines · 35 tokens per session scan A 0ab7edf3fc37

Subscribe to this mod's changes

rescue is a command published in the GitHub repository limeflash/antigravity-plugin-cc (2 stars, last pushed 1mo ago), licensed MIT. It adds 35 tokens to every session and 1,313 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.