implement

A command that asks Gemini, through the Antigravity CLI, to implement a clearly specified coding task by editing files directly. It first checks that the work is inside a Git repository so the changes can be reviewed.

In plain words
What is it for?
It helps with batch renames, boilerplate, test scaffolding, and applying one established code pattern across multiple files.
Why use it?
It handles repetitive, well-defined edits while preserving a before-and-after record of the files changed.

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/haunchen/gemini-plugin-cc/implement
Clone the repo
git clone --depth 1 https://github.com/haunchen/gemini-plugin-cc
Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,482 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 $0.00014 $0.01482
Opus 5 $0.00007 $0.00741
Sonnet 5 $0.00003 $0.00296
Haiku 4.5 $0.00001 $0.00148

Measured yesterday against content hash d09a30032e9b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

implement 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/gemini/commands/implement.md · 108 lines

How it starts

The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Implement a task using Gemini via the Antigravity CLI (agy). Unlike /gemini:review, this command writes to your files.

Use it for work that is mechanical, well-specified, and cheap to verify — batch renames, boilerplate, test scaffolding, applying one pattern across several files. Work that needs whole-repo judgment is better done in this session, where the context already exists.

Step 1: Establish the workspace

The agent writes files, so a non-recoverable run is not acceptable. Get the repository root:

ROOT=$(git rev-parse --show-toplevel 2>/dev/null)

If this fails, stop. Tell the user /gemini:implement only runs inside a git repository, because git diff is the only thing that makes the agent's writes reviewable and reversible. Do not offer to run it anyway.

Then capture what the tree looks like before the run — this is what makes an unannounced write visible in Step 5:

git status --porcelain

Keep the output as BEFORE. If it is non-empty, tell the user which files are already dirty and ask whether to continue. Uncommitted work is not a blocker, but after the run there is no way to tell their edits from Gemini's, and that is worth one question. If they decline, stop.

Step 2: Parse arguments

--model <value>: extract as MODEL and remove the pair from $ARGUMENTS. If absent, MODEL = gemini-3.7-flash-high.

Aliases: flashgemini-3.7-flash-high, progemini-3.1-pro-high. Anything else passes through unchanged (agy models lists the slugs).

The default is 3.7 rather than the 3.6 that /gemini:review uses, and the split is deliberate: measured on this repo's eval suite the two are indistinguishable at reviewing, so review stays where it is, while 3.7's gains are in writing code, which is this command's entire job.

--brief <path>: Read the file and use its contents as BRIEF. Remove the pair from $ARGUMENTS. If the path does not exist, say which one and stop.

--context <path> (repeatable): for each, Read the file (expand with Glob first if it contains * or ?). Concatenate the contents, each preceded by a line naming the file, as CONTEXT. Remove every --context <value> pair from $ARGUMENTS. A missing path stops the command — a silently dropped context file produces a plausible implementation built on nothing.

Read the full file on GitHub · 108 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 · 108 lines · 14 tokens per session scan A d09a30032e9b

Subscribe to this mod's changes

implement is a command published in the GitHub repository haunchen/gemini-plugin-cc (1 stars, last pushed 18d ago), licensed MIT. It adds 14 tokens to every session and 1,482 once invoked, about $0.0001 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.