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.
npx agentmods add commands/bytelandtechnology/gemini-cc/setupgit clone --depth 1 https://github.com/ByteLandTechnology/gemini-ccWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00013 | $0.01229 |
| Opus 5 | $0.00006 | $0.00615 |
| Sonnet 5 | $0.00003 | $0.00246 |
| Haiku 4.5 | $0.00001 | $0.00123 |
Grade A, and why
setup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
First, run the setup helper with any explicit flags the user provided:
node "${CLAUDE_PLUGIN_ROOT}/scripts/gemini-companion.mjs" setup --json $ARGUMENTS
Parse the JSON output to extract:
ready— whether Gemini is installed and authenticatedprovider.available— whether Gemini CLI is installedauth.loggedIn— whether Gemini is authenticatedauth.detail— authentication detail messagereviewGateEnabled— current review gate statedefaultModel— current default model (may be null)
If explicit flags were provided (i.e., $ARGUMENTS is not empty):
- Present the final setup output to the user.
- Stop here.
If no explicit flags were provided, proceed to interactive mode:
If provider.available is false and npm is available:
- Use
AskUserQuestionexactly once to ask whether Claude should install Gemini now. - Put the install option first and suffix it with
(Recommended). - Use these two options:
Install Gemini CLI (Recommended)Skip for now
- If the user chooses install, run:
npm install -g @google/gemini-cli
- Then rerun:
node "${CLAUDE_PLUGIN_ROOT}/scripts/gemini-companion.mjs" setup --json
If provider.available is true but auth.loggedIn is false:
- Present the auth guidance from
auth.detailand stop. - Do not offer to install; the CLI is already installed.
- Preserve guidance to run
geminiand complete/auth, or setGEMINI_API_KEY/GOOGLE_API_KEY.
If ready is true, present a menu using AskUserQuestion to configure settings.
Menu questions (ask one at a time in this order):
-
Model selection — ask which model to use as default:
question: "Which model should be the default for /gemini:rescue and /gemini:review?" header: "Default Model" options: - label: "gemini-3.1-pro-preview (Recommended)" description: "Latest pro, highest capability" - label: "gemini-2.5-pro" description: "Pro tier, strong reasoning" - label: "gemini-3-flash-preview" description: "Latest flash, fast and capable" - label: "gemini-2.5-flash" description: "Fast, lower cost, good for most tasks" - label: "gemini-3.1-flash-lite-preview" description: "Latest lite model, lowest latency" - label: "gemini-2.5-flash-lite" description: "Lighter weight, lowest cost" - label: "Auto" description: "Let Gemini CLI decide" - label: "Custom model..." description: "Enter a custom model name" multiSelect: false- If user selects "Auto", run:
node "${CLAUDE_PLUGIN_ROOT}/scripts/gemini-companion.mjs" setup --set-model "" --json - If user selects "Custom model...", ask a follow-up question:
The user's answer is the custom model name. Run:question: "Enter the custom model name:" header: "Custom Model" options: - label: "Continue" description: "Use the model name you provide" multiSelect: falsenode "${CLAUDE_PLUGIN_ROOT}/scripts/gemini-companion.mjs" setup --set-model <answer> --json - Otherwise run:
node "${CLAUDE_PLUGIN_ROOT}/scripts/gemini-companion.mjs" setup --set-model <model-id> --json(Use the model ID from the selected option label, e.g.gemini-2.5-flash)
- If user selects "Auto", run:
-
Review gate — if
reviewGateEnabledis false, ask:question: "Enable the stop-time review gate? When enabled, Gemini reviews Claude's response before each stop and may block the stop if issues are found." header: "Review Gate" options: - label: "Enable review gate" description: "Require Gemini review before stops" - label: "Keep disabled (Recommended)" description: "No extra review at stop time" multiSelect: falseIf
reviewGateEnabledis true, ask:question: "The review gate is currently enabled. Disable it?" header: "Review Gate" options: - label: "Disable review gate" description: "Stop requiring Gemini review before stops" - label: "Keep enabled" description: "Continue requiring review before stops" multiSelect: false
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.
- 2d ago First seen · 133 lines · 13 tokens per session scan A 4af5a363dde1
setup is a command published in the GitHub repository ByteLandTechnology/gemini-cc (1 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 13 tokens to every session and 1,229 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.
Other commands, from other repositories
frontend
Injects context of all relevant cli files.
derive-tests
Turn documented intent into a test-coverage map — inventory the tests that exist today, derive use-case cases from the system docs, separate existing coverage from proposed tests and unverified gaps, mark each unit / guarded-live / manual, and recommend a green-before-merge CI gate.
plan-launch
Create a full go-to-market strategy — beachhead segment, ICP, messaging, channels, and launch plan.
competitive-analysis
Analyze the competitive landscape — identify competitors, compare strengths and weaknesses, find differentiation opportunities.
spec-design
Create comprehensive technical design for a specification.
OPSX: Bulk Archive
Archive multiple completed changes at once.