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/haunchen/gemini-plugin-cc/setupgit clone --depth 1 https://github.com/haunchen/gemini-plugin-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 | $0.00021 | $0.01159 |
| Opus 5 | $0.00010 | $0.00580 |
| Sonnet 5 | $0.00004 | $0.00232 |
| Haiku 4.5 | $0.00002 | $0.00116 |
Grade B, and why
setup scanned grade B with 1 finding 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
Expect `agents : 4 processed` in the output. This installs four agents into `~/.gemini/config/plugins/gemini-agents/`: How it starts
The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Step 0: Parse --model parameter
Check if $ARGUMENTS contains --model <value>:
- If yes: extract the value as MODEL, remove
--model <value>from $ARGUMENTS - If no: set MODEL =
gemini-3.6-flash-high
Run the following checks in order and report the results.
1. Check if agy is installed
Run: which agy || where agy 2>/dev/null
- If found: report the path
- If not found: tell the user to install the Antigravity CLI from https://antigravity.google and stop here
2. Check agy version
Run: agy --version
Markdown custom agents require 1.1.6 or newer — that is how this plugin injects its system prompts. If the version is older, tell the user to run agy update and stop here.
3. Check authentication
Run: agy -p "ping" --model $MODEL --print-timeout 2m 2>&1 | head -5
- If it returns a normal reply: authentication is working
- If it reports an authentication or eligibility error: tell the user to run
agyinteractively once to sign in via Google OAuth
4. Install the agents
Determine the absolute path to the plugin root (the parent of the commands/ directory containing this file). The agy plugin lives in agy/ under that root.
Run: agy plugin install "<plugin-root>/agy"
Expect agents : 4 processed in the output. This installs four agents into ~/.gemini/config/plugins/gemini-agents/:
gemini-review— used by/gemini:review(read-only)gemini-adversarial-review— used by/gemini:adversarial-review(read-only)gemini-ask— used by/gemini:ask(read-only)gemini-implement— used by/gemini:implement(writes files)
The first three carry a read-only whitelist. gemini-implement is the exception: it can create and edit files, which is the whole point of that command. It still has no shell.
Re-running this command upgrades an existing install in place.
5. Verify the agents actually took effect
This step is not optional. agy --agent <name> silently ignores unknown agent names — a failed install produces plausible-looking output with none of the plugin's system prompt applied, and the user would never notice.
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.
- yesterday First seen · 97 lines · 21 tokens per session scan B d6828db78cfc
setup is a command published in the GitHub repository haunchen/gemini-plugin-cc (1 stars, last pushed 19d ago), licensed MIT. It adds 21 tokens to every session and 1,159 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
package-assistant
Convert an OfficeCLI skill into a fully wired AionUi assistant preset, or update an existing one.
spec-design
Create comprehensive technical design for a specification.
spec-tasks
Generate implementation tasks for a specification.
spec-requirements
Generate comprehensive requirements for a specification.
validate-gap
Analyze implementation gap between requirements and existing codebase.
spec-impl
Execute spec tasks using TDD methodology.