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 skills/monarchjuno/tradingcodex/tcx-calculationnpx skills add monarchjuno/tradingcodex --skill tcx-calculationgit clone --depth 1 https://github.com/monarchjuno/tradingcodexWhat 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 | $0.00059 | $0.00801 |
| Opus 5 | $0.00030 | $0.00400 |
| Sonnet 5 | $0.00012 | $0.00160 |
| Haiku 4.5 | $0.00006 | $0.00080 |
Grade A, and why
tcx-calculation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reproducible Financial Calculation
Use the smallest governed input and leave an auditable Calculation Run whenever the result can affect a conclusion. Keep quick arithmetic that does not support a conclusion explicitly exploratory.
Follow the calculation workflow
- Search Dataset and Calculation cards before fetching data or computing.
- Inspect only the relevant manifest or run summary. Confirm source lineage,
units, currency, timezone, adjustment policy, and
knowledge_cutoff. - Reuse a prior result only when
prepare_calculationreports an exact fingerprint match. Treat similar runs as references, not cached answers. - Materialize only the needed columns, instruments, and time range. Keep private portfolio or ledger inputs run-scoped; never register them as a Dataset or copy them into scripts, logs, or artifacts.
- Create one direct basename-only
.pyfile under$TRADINGCODEX_SCRATCHwith nativeapply_patch. For a conclusion-relevant calculation, callprepare_calculationbefore execution and use only its declared inputs and outputs. - Run exactly
./tcx-calc <filename.py>from the workspace root on POSIX or.\\tcx-calc.cmd <filename.py>on Windows. Do not invoke system Python, install packages, use heredocs, or pass-c,-m, paths, or extra args. - Call the runner-injected
tcx_emit_resultglobal exactly once with one positional object. Never import it, pass keyword arguments, invent wrapper fields, or emit a metrics mapping. Copy the exact typed shape from references/data-runtime.md. - Record success or failure with
record_calculation_run. On failure, read its safeerror_codeanderror_message, make one concrete correction, stage a new script basename, prepare a new immutable spec, and retry. Never overwrite a prepared script/result, repeat the same failed code unchanged, install packages, or reuse a failed Run. Stop and hand off aswaitingif the same error code recurs after its targeted correction. - Bind accepted
calculation_run_ids, Dataset lineage, assumptions, diagnostics, and warnings into the role artifact. Do not cite an exploratory sidecar-free execution as decision evidence.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 65 lines · 59 tokens per session scan A 3cc7c36317ed
tcx-calculation is a skill published in the GitHub repository monarchjuno/tradingcodex (364 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 59 tokens to every session and 801 once invoked, about $0.0003 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 skills, from other repositories
agent-agentic-payments
Agent skill for agentic-payments - invoke with $agent-agentic-payments.
agent-payments
Agent skill for payments - invoke with $agent-payments.
company-research
A 股个股研究六阶段 SOP(profile → financials → estimates → valuation → risk → report),Phase 0 范围 = 财务估值闭环。当任务是研究 / 分析 / 评估一只或多只已指定代码的 A 股个股时使用;规定每阶段取哪些数据、调哪些 calc 函数、必须落盘什么产物、过什么 Gate。不用于:从市场中筛选标的、泛行业讨论、概念解释、给投资动作建议。.
data-access
A 股零鉴权取数手册。当需要真实的行情 / 市值 / 估值快照、季度报告期累计财务数据、机构一致预期 EPS、PE 历史序列、公告标题、日 K 线、交易日历时使用;只允许运行本 skill 登记的脚本取数(腾讯 / 新浪 / 同花顺 / baostock / 深交所 / 东财),禁止凭模型记忆给数,禁止自造爬虫。概念解释、观点讨论等不需要取数的话题不要加载。.
industry-chain
产业链下钻与不可替代性判定方法:以龙头为"需求入口"沿供应链逐层下钻(整机 / 龙头 → 部件 → 核心器件 → 材料 → 衬底与设备),用物理 / 材料约束(扩产周期、良率、认证周期、有无替代)当筛子找供给刚性的卡口;给每个标的贴不可替代性标签(techmoat / capacitymoat / both / 待补)并列证据;含"卡口越硬越贵"与预期差四问的校准。当任务涉及产业链位置、上下游、护城河、不可替代性、供给瓶颈、竞争格局时加载;单纯取数、估值计算、财报拆分等不涉及产业链结构的任务不要加载。只产出框架与证据表,不给投资动作建议。.
catalyst-risk
催化剂与风险的反证式写法:每个强结论必须先找反证;催化剂按"兑现型 / 预期型 / 周期型"分类并要求可验证的数据时点;风险按技术路线断层、客户集中、产能过剩与价格战、周期顶、预期透支(假便宜 PEG)、一致预期下修、治理与流动性、数据源冲突分类;裁决点的标准写法(什么数据出来会改变判断 + 下一个公开数据时点);知识档案旧结论的反证处理。当任务涉及风险、反证、催化剂、裁决点、预期兑现、什么会推翻结论时加载;单纯取数、估值计算、财报拆分等不需要反证框架的任务不要加载。只产出框架、概率与裁决点,不给投资动作建议。.