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/singleflo/odoo-assistant-mcp/agents-mdgit clone --depth 1 https://github.com/singleflo/odoo-assistant-mcpWrote 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/singleflo/odoo-assistant-mcp/agents-md)<a href="https://agentmods.dev/instructions/singleflo/odoo-assistant-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/singleflo/odoo-assistant-mcp/agents-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 | $0.02017 | $0.02017 |
| Opus 5 | $0.01009 | $0.01009 |
| Sonnet 5 | $0.00403 | $0.00403 |
| Haiku 4.5 | $0.00202 | $0.00202 |
Grade A, and why
odoo-assistant-mcp AGENTS.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 3d 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — odoo-assistant-mcp
MCP server (stdio) that wraps 9 verified Odoo scripts as 19 tools + 8
resources. Shipped and complete; published on PyPI as version 0.1.1 and
installable with uvx odoo-assistant.
Read before coding
docs/PRD.md— the spec: tools (§5B), auth (§6), safety mapping (§7), packaging (§10). Written before the build, so it trails the code: it says 14 tools and two sections are marked superseded. Trust the code first.docs/mcp-spec/— the ENTIRE MCP 2026-07-28 spec, vendored (144 files). Consult it instead of the web.docs/RELEASE-CHECKLIST.md— the only work left, and all of it needs a human: PyPI Trusted Publishing, then the tag.docs/json2-source/— the Odoo module behind/json/2.odoo_client.pyprobes that transport first and falls back to XML-RPC, which Odoo 19 deprecates and 22 removes, so this is the reference for the path that becomes primary.references/BUILD-STATE.md— sanity figures for the dev instance
Commands
uv run pytest # 225 tests. addopts already excludes live+wheel
uv build && uv run pytest -m wheel # 3 tests. needs dist/ AND network (uvx)
uv run pytest -m live # 6 tests. needs the env vars below
Never write -m "not live". A -m on the command line replaces addopts
instead of narrowing it, so that spelling silently re-enables the wheel tests
and they fail on a clean checkout with no dist/. CI spells it in full:
-m "not live and not wheel".
Protocol check: npx @modelcontextprotocol/inspector against the server.
Hard rules
- Never rewrite
src/odoo_assistant/odoo_scripts/. Those 9 scripts are verified against a live instance and are the source of truth; the server is a thin wrapper importing them. They came from an agent skill that has been retired, so this repo is the canonical copy —scripts/sync-from-skill.shhas no upstream left and exits 1. Fix a bug there only with a live test proving it; otherwise adapt the wrapper. - Scripts are flat modules: each does
sys.path.insert(0, <own dir>)+from odoo_client import .... Keep that import style working when packaging. - SDK:
from mcp.server import MCPServer(SDK 2.0, pinnedmcp[cli]>=2.0.0,<3).MCPServer(...)defaultsversionto""— pass it explicitly or every host displays an empty version. - Module decomposition:
server.pystays thin. Tools and resources live intools_read.py/tools_write.py/tools_collab.py/tools_discuss.py/tools_evolution.py/resources.py, each exposingregister(mcp), wired byserver._register_all(). Adding a tool means updating the three tests that assert the exact tool set.tools_collab.pynotifies ABOUT a record (Inbox bell);tools_discuss.pyis user-to-user conversation (chat systray) — "message X" is the second, and itschannel_getis the one method the verifiedsafety_layer.pywas extended to admit (L1). explore_module.REF_DIRredirect needs the package-qualified import (from odoo_assistant.odoo_scripts import explore_module) so the patch is visible to every importer. Seetools_evolution.py's docstring.- Wheel bundles
references_public/only (scrubbed, generic).references/holds the instance-specific set and is excluded. - stdio transport: nothing may reach stdout but the JSON-RPC stream. Diagnostics go to stderr.
- Every write passes
server_safety.gate()(L0–L5). L4/L5 refuse by default;ODOO_MCP_MAX_LEVELmoves the bar and refuses startup if invalid. Host consent dialogs are untrusted — the gate is the only enforcement point.
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.
- 3d ago First seen · 83 lines · 2,017 tokens per session scan A d58fdafe4eef
odoo-assistant-mcp AGENTS.md is an instructions file published in the GitHub repository singleflo/odoo-assistant-mcp (0 stars, last pushed 9d ago), licensed MIT. It adds 2,017 tokens to every session, about $0.0101 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
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.