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 instructions/actinon-com/brass-monkey/gemini-mdgit clone --depth 1 https://github.com/actinon-com/brass-monkeyWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/actinon-com/brass-monkey/gemini-md)<a href="https://agentmods.dev/instructions/actinon-com/brass-monkey/gemini-md"><img src="https://agentmods.dev/badge/instructions/actinon-com/brass-monkey/gemini-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01602 | $0.01602 |
| Opus 5 | $0.00801 | $0.00801 |
| Sonnet 5 | $0.00320 | $0.00320 |
| Haiku 4.5 | $0.00160 | $0.00160 |
Grade A, and why
brass-monkey GEMINI.md 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 4d 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Brass-Monkey (Gemini CLI Extension for Odoo)
Project Mandate
Brass-Monkey replaces Brass-Compass, a prototype Python MCP server, with a modern, secure, and intelligent Node.js bridge for Odoo. It provides AI agents with a high-fidelity interface to Odoo's CRM, ERP, and business logic layers.
Source Reference: Original Python implementation: brass-compass
Critical Protocols
1. Orientation (The "World Map" Rule)
- Mandate: Your ABSOLUTE FIRST ACTION when starting a new session or connecting to an Odoo instance for the first time MUST be:
- Activate the
odoo-get-startedskill. - Execute
get_environmentto establish a context "World Map".
- Activate the
- Reasoning: Odoo instances vary wildly. "Blind" discovery via multiple
search_readcalls is inefficient and error-prone. Oneget_environmentcall provides server version, user permissions, active company, and installed apps. - Enforcement: Failure to orient before executing business logic is considered a violation of project standards.
2. Explanations First & Buy-in
- Mandate: ALWAYS provide a technical explanation of proposed changes and seek user understanding/approval BEFORE modifying files or executing state-changing commands.
- UI Transparency: When using state-changing tools (
create_record,write_record,unlink_record), you MUST describe the intended change and the specific values in the chat message before calling the tool. Because tool parameters are often truncated in the UI, the chat message is the primary source of truth for user approval. - Enforcement: For any change affecting more than one file, modifying tool logic, or altering architectural patterns, you MUST present a strategy and wait for explicit user buy-in. Treat reports of bugs or concerning behavior as Inquiries (Research/Strategy) rather than Directives (Immediate Action) until a plan is agreed upon.
4. Structural Fixes over Band-aids
- Mandate: Prioritize correcting the root cause of agent confusion over implementing restrictive "Write Guards" or tool-level blocks.
- Pattern: If an agent makes a mistake, first audit the Orientation (World Map) data provided by
get_environmentand the Skill instructions. Fix the data presentation or the guidance before resorting to hard-coded tool restrictions.
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.
- 4d ago First seen · 102 lines · 1,602 tokens per session scan A fff45c4d273b
brass-monkey GEMINI.md is an instructions file published in the GitHub repository actinon-com/brass-monkey (2 stars, last pushed 11d ago), licensed MIT. It adds 1,602 tokens to every session, about $0.0080 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 instructions, from other repositories
orbiteus AGENTS.md
Instructions for orbiteus/orbiteus, covering agent / contributor policy, vendor-neutral copy and links, readme hero tagline (locked), stack and boring tech filter.
S.HiveDesk AGENTS.md
AGENTS.md instructions for Akshit1018/S.HiveDesk, covering app builder workspace, 0. two worlds (read this first), project instructions, where you are and /workspace/startup.sh (required — you maintain this).
able AGENTS.md
AGENTS.md instructions for codeyogi911/able, covering able, goal, working context, operating model and scope and autonomy.
gemini-watch GEMINI.md
Gemini CLI instructions for sudoax0n/gemini-watch, covering gemini watch extension, step 0 — setup preflight (runs every invocation), step 1 — parse input, step 2 — run the watch script and useful flags.
able CLAUDE.md
Claude Code instructions for codeyogi911/able, a project described as: Agent-first, open-source ERP for service businesses. Able Desk — a support desk and CRM on Cloudflare Workers, operated through MCP — is the first application.
gemini-cli-desktop AGENTS.md
Instructions for Piebald-AI/gemini-cli-desktop, a project described as: Web/desktop UI for Gemini CLI/Qwen Code. Manage projects, switch between tools, search across past conversations, and manage MCP servers, all from one multilingual interface, locally or remotely.