Borrowing it
Nothing to install: this file belongs to up2itnow0822/ReadyTrader-FOREX. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/up2itnow0822/ReadyTrader-FOREX/main/AGENTS.mdgit clone --depth 1 https://github.com/up2itnow0822/ReadyTrader-FOREXWrote 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/up2itnow0822/readytrader-forex/agents-md)<a href="https://agentmods.dev/instructions/up2itnow0822/readytrader-forex/agents-md"><img src="https://agentmods.dev/badge/instructions/up2itnow0822/readytrader-forex/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/up2itnow0822/readytrader-forex/agents-md"><img src="https://agentmods.dev/badge/instructions/up2itnow0822/readytrader-forex/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.01479 | $0.01479 |
| Opus 5.5 | $0.00592 | $0.00592 |
| Sonnet 5.5 | $0.00296 | $0.00296 |
| Haiku 4.5 | $0.00148 | $0.00148 |
Grade A, and why
ReadyTrader-FOREX 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 8d 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.
ReadyTrader-FOREX
Root AGENTS.md (DOX rail).
Purpose
A stdio MCP server (app/main.py, 29 tools) that lets an AI agent research currency pairs, trade a
persistent paper FX account, and optionally place live orders through OANDA behind a Risk Guardian,
a price-based market guard, an operator policy and a kill switch; plus an approval/dashboard API
(app/api_server.py) and a Next.js dashboard (frontend/).
Ownership
app/: MCP tool surface (app/tools/), settings (app/core/config.py), the container, the approval API.core/: pure rules and state: risk (risk.py), Falling Knife and volatility halt (market_guard.py), policy, the paper FX account (fx_account.py), backtest.execution/: brokerage connectors (OANDA is the only FX venue) and the approval store;marketdata/,intelligence/: data, calendar and news sources.common/: shared helpers (switches.py,paths.py).research/falling_knife/: the study behinddocs/FALLING_KNIFE.md; nothing imports it._deprecated/: history only, never loaded or shipped (see its README).configs/: copy/paste MCP client configs, each in its client's own format (Agent Zero:agent_zero.mcp.json, the{"mcpServers": ...}JSON for Settings → MCP/A2A → External MCP Servers);tests/test_configs.pykeeps them and the README's examples in step.
Local Contracts
- Fail closed. Safety switches are parsed with
common/switches.py: protections stay on unless explicitlyfalse/0/no/off;LIVE_TRADING_ENABLEDneeds exactlytrue; the kill switch halts on any value but empty/off; an unknownEXECUTION_APPROVAL_MODErequires approval; an unreadable limit refuses live orders. Brokerage connectors default to practice/paper/sandbox accounts. - Every live order passes
pre_trade_checkandlive_order_refusalwhen placed or proposed, and again at execution (including/api/approve-trade). No simulated venue is registered for live orders, and a proposal executes only in the mode (paper/live) it was made in. - FX positions go both ways: risk rules judge the exposure an order adds (from the paper account or the brokerage's positions), not its side. Size = USD notional of the added part, valued at the market (never at a price the caller supplies alone; a limit at the higher of limit and market); an order that adds exposure and cannot be valued or sized is refused; reducing or closing a position is never sized as new exposure. Unknown live positions count as new exposure.
- Numeric MCP tool parameters use
app/tools/params.py(Number,Integer): booleans are refused. Quotes are used only when finite and positive. In live mode the operator switches answer first, in the order tools and at approval; every well-formed order request is audited (trade_start). - Symbols are normalised (
app/tools/trading.canonical_symbol: any pair spelling ->EURUSD) before any check, so every spelling gets the same market checks. - Nothing the server sends rests at OANDA: every order is fill-or-kill (a limit is a MARKET order
with
priceBound),positionFill: REDUCE_FIRST. The Guardian sizes against current positions, never open orders. - Paper loss limits are a time-weighted index over
fx_equitysnapshots (deposits are neither gains nor losses);mark_day_openrecords the day's baseline at the first check; a position that cannot be priced makes equityNoneand adding orders fail closed. Live accounts have no loss history: live verdicts listdaily_loss_limit/max_drawdownunderinactive_rules. - The approval API:
API_OPERATOR_TOKEN(when set) guards every/api/route but/api/healththrough therequire_operatordependency of theoperator_apirouter (never a path check in a middleware: the URL path can differ from the routed path); new protected routes go onoperator_api. CORS is added after therequest_contextmiddleware so it wraps every answer; a live proposal is approved only when it is set (checked before the proposal is consumed). Every response carriesX-Request-ID; errors never echo exception text. - Paper orders fill only in
core/fx_account.FxPaperAccount(shared by the MCP server and the API throughdata/paper.db). - Tools answer
{"ok": true, "data"}or{"ok": false, "error": {"code", "message", "data"}}; a source that cannot answer is an error, never a payload (never "no events"). New codes go indocs/ERRORS.md. - A price is finite and positive or it is no price: provider rows without one are dropped
(
marketdata/exchange_provider._priced_rows), quote tools answer only real prices (_real_price), and the paper account refuses a non-finite rate (NaN is truthy and compares false against every limit). - Default data files live in
<repo>/data/viacommon/paths.data_path(never the working directory);READYTRADER_DATA_DIRand each*_PATHvariable override. - Every variable the code reads is in
env.example, with no inline comments or placeholder keys. - Docs describe only what runs: after changing a tool, run
python tools/generate_tool_docs.py(it reads the live registry) and update README /docs/in the same change.
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.
- 8d ago First seen · 83 lines · 1,479 tokens per session scan A 3a1b8b941339
ReadyTrader-FOREX AGENTS.md is an instructions file published in the GitHub repository up2itnow0822/ReadyTrader-FOREX (0 stars, last pushed 9d ago), licensed MIT. It adds 1,479 tokens to every session, about $0.0059 per session on Opus 5.5. 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 2026-10-12.
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).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
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 copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.