Menu
Reference for a disabled Gemini CLI sandbox capability.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Search Zepto, manage the shared account and cart, and place explicitly confirmed final orders for the authorized private Telegram chat.
Expertise in google-adk Agent framework. Use when the user asks for feature/ changes in their code related to google-adk
Core agent-browser usage guide. Read this before running any agent-browser commands. Covers the snapshot-and-ref workflow, navigating pages, interacting with elements (click, fill, type, select), extracting text and data, taking screenshots, managing tabs, handling forms and auth, waiting for content, running multiple browser sessions in parallel, and troubleshooting common failures. Use when the user asks to interact with a website, fill a form, click something, extract data, take a screenshot, log into a site, test a web app, or automate any browser task.
Clone and explore GitHub repositories using sandbox tools. Understand codebase structure, read key files, and answer questions about any public or private repository.
| name | gemini_cli |
| description | Reference for a disabled Gemini CLI sandbox capability. |
The Gemini CLI (@google/gemini-cli) is a command-line interface for interacting with the Google Gemini API.
Blacki does not register this skill. General-purpose sandboxes intentionally
receive no model or repository credentials, so enabling the commands below would
create a non-functional or unsafe capability.
Do not enable this skill by injecting a standing API key. A future implementation must use an explicit, least-privilege capability broker with short-lived credentials and separate allow/deny tests.
For reference, the Gemini CLI installation command is:
hash gemini 2>/dev/null || npm install -g @google/gemini-cli@latest
Since you are an automated agent, you MUST use the Gemini CLI in non-interactive mode.
Use the -p (or --prompt) flag to send a query non-interactively. This will execute the task and return the output without dropping into the REPL.
sandbox_run_command(command='gemini -p "Summarize the README.md file in the current directory."')
You can pipe file contents into Gemini for context, or rely on its ability to read the workspace.
sandbox_run_command(command='cat main.py | gemini -p "Explain this code and suggest refactoring."')
You can resume a previous session if you need to continue a conversation.
sandbox_run_command(command='gemini -r "latest" "Apply the suggested refactoring to the code."')
This page provides a reference for commonly used Gemini CLI commands, options, and parameters.
| Command | Description | Example |
|---|---|---|
gemini -p "query" | Query non-interactively | gemini -p "summarize README.md" |
cat file | gemini | Process piped content | cat logs.txt | geminiGet-Content logs.txt | gemini |
gemini -r "latest" | Continue most recent session | gemini -r "latest" |
gemini -r "latest" "query" | Continue session with a new prompt | gemini -r "latest" "Check for type errors" |
gemini update | Update to latest version | gemini update |
| Argument | Type | Description |
|---|---|---|
query | string (variadic) | Positional prompt. Defaults to interactive mode in a TTY. Use -p/--prompt for non-interactive execution. |
| Option | Alias | Type | Default | Description |
|---|---|---|---|---|
--model | -m | string | auto | Model to use. See Model Selection for available values. |
--prompt | -p | string | - | Prompt text. Appended to stdin input if provided. Forces non-interactive mode. |
--resume | -r | string | - | Resume a previous session. Use "latest" for most recent or index number (for example --resume 5) |
--output-format | -o | string | text | The format of the CLI output. Choices: text, json, stream-json |