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/armandcismaru/deeptraderx/claude-mdgit clone --depth 1 https://github.com/armandcismaru/DeepTraderXWrote 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/armandcismaru/deeptraderx/claude-md)<a href="https://agentmods.dev/instructions/armandcismaru/deeptraderx/claude-md"><img src="https://agentmods.dev/badge/instructions/armandcismaru/deeptraderx/claude-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 5.1 | $0.00620 | $0.00620 |
| Opus 5 | $0.00310 | $0.00310 |
| Sonnet 5 | $0.00124 | $0.00124 |
| Haiku 4.5 | $0.00062 | $0.00062 |
Grade A, and why
DeepTraderX CLAUDE.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 5d 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Guidance for Claude Code when working in this repository. The full guidance —
conventions, the concurrency model, security findings, and the catalogue of
known bugs — lives in AGENTS.md and is imported below. Read it before editing.
@AGENTS.md
Quick reference
What it is: DeepTraderX — an LSTM trading agent (DTR) competing inside
TBSE, a multi-threaded simulated limit-order-book market. Academic research code;
no live trading, no network surface. Full docs in docs/.
Run it (always from deep_trader_tbse/):
pip install -r requirements.txt
cd deep_trader_tbse
python3 tbse.py # config-driven
python3 tbse.py 0 0 0 5 0 0 5 # 7 ints: ZIC ZIP GDX AA GVWY SHVR DTR per side
python3 tbse.py markets.csv # batch; prints "Done Now" on success
Verify / check: no test suite. Success = run completes (CSV mode prints
Done Now) and pylint/black pass (./checks.sh).
Things to remember (high-signal)
- Run from
deep_trader_tbse/— imports and the model path are CWD-relative. - Flaky trials are normal. Sessions ending with the wrong thread count are
auto-discarded and retried. The shared
Exchangeis unsynchronized; the GIL + retry loop are load-bearing. Don't add unlocked compound reads overtape/lob_anon. docs/is published to public GitHub Pages — keep security findings and internal notes out of it (they belong here / inAGENTS.md).- Bug catalogue is in
AGENTS.md§6. The high-severity correctness bugs (Shaver/Sniper selling below cost, single-trader divide-by-zero,timemodemismatch, broken training entry point,pickle_files/bookkeep/DataGeneratordefects) were fixed on 2026-06-02. Still open: dead/drifting deployment code (S3 calls commented out, Dockerfile not usingrequirements.txt, macOS arm64 TF/Keras version risk). - Keep features in sync:
Exchange.lob_data_out()(training data) andDeepTrader.create_input()(inference) must produce the same 13 features in the same order. - Style: black + permissive pylint (
.pylintrc, max line 120). Preserve the ported BSE/TBSE structure; prefer minimal surgical diffs. - Commit/push only when asked. Default branch is
main.
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.
- 5d ago First seen · 48 lines · 620 tokens per session scan A fa6c71717a19
DeepTraderX CLAUDE.md is an instructions file published in the GitHub repository armandcismaru/DeepTraderX (21 stars, last pushed 19d ago), licensed MIT. It adds 620 tokens to every session, about $0.0031 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 2026-08-30.
Other instructions, from other repositories
albucore AGENTS.md
Instructions for albumentations-team/albucore, covering codex ai development guidelines for albucore, project overview, core documentation, repo-local codex skills and development principles.
ApraPipes CLAUDE.md
Claude Code instructions for Apra-Labs/ApraPipes, covering claude.md - aprapipes declarative pipeline, current phase: sprint 12 - windows integration test fix, sdk structure (complete), critical rules and 1. build and test before commit (mandatory).
zero-to-ai snyk_rules.instructions.md
Snyk Security At Inception.
DeepSpeed AGENTS.md
AGENTS.md instructions for deepspeedai/DeepSpeed, covering deepspeed project rules, commit & ci requirements, code change discipline, test discipline and tool caveats.
ai AGENTS.md
AGENTS.md instructions for vercel/ai, covering agents.md, project overview, repository structure, key directories and core package dependencies.
ai CLAUDE.md
Claude Code instructions for vercel/ai, a project described as: The AI Toolkit for TypeScript. From the creators of Next.js, the AI SDK is a free open-source library for building AI-powered applications and agents.