agentleFS
BryceYuuuBryceYuuu/open-workflow-skills/skills/strategy-backtest/SKILL.md

strategy-backtest

Skill1 starsChanged 2 months agoMITView on GitHub

Convert explicit trading rules into historical backtests with fees, slippage, and bias checks; never execute live trades by default. Use when the user asks for strategy backtest or an equivalent workflow.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 13 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈13

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when the user explicitly requests strategy backtest, or when the requested outcome clearly matches this workflow.

    line 45
  2. Distinctive:

    Do not activate it merely because the topic is related.

    line 47
  3. Distinctive:

    Treat permissions as capabilities to request or verify, not as permissions automatically granted by installing the skill.

    line 68
  4. Distinctive:

    No universal runtime package is required by the skill format itself.

    line 82
  5. Distinctive:

    Verify all expected outputs exist and are readable.

    line 122
  6. Distinctive:

    Verify the source was not unintentionally overwritten.

    line 123
  7. Distinctive:

    Verify structured outputs against their expected schema or invariants.

    line 124
  8. Distinctive:

    Never mark the workflow successful solely because a model produced text.

    line 126
  9. Distinctive:

    Do not invent source facts or claim an external action succeeded without verification.

    line 132
  10. Distinctive:

    Treat outputs as analysis, not guaranteed financial outcomes.

    line 135
  11. Distinctive:

    If a dependency, credential, connected service, browser session, file, or required permission is unavailable:

    line 140
  12. Distinctive:

    Stop the affected step.

    line 142
  13. Distinctive:

    Do not fabricate a successful result.

    line 145

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Strategy Backtest / 交易策略回测

Purpose / 用途

中文

把明确交易规则转换为历史回测,纳入手续费、滑点和偏差检查;默认绝不执行真实交易。

English

Convert explicit trading rules into historical backtests with fees, slippage, and bias checks; never execute live trades by default.

Status / 状态

  • Workflow status: 🟡 Experimental / Assisted / 实验性 / 辅助型 / 辅助完成
  • Implementation: 📘 Definition only / 仅工作流定义
  • Audit grade: C
  • Quality score: 3.5/5
  • Automation level: L2-L3
  • Category: Business & Growth / 商业与增长

Workflow status describes how well the workflow itself is understood. Implementation status separately tells you whether this repository ships runnable code for it. Integration skills still require external services or authorization; experimental skills may need more human review.

Search aliases / 搜索关键词

strategy-backtest · Strategy Backtest · 交易策略回测 · strategy backtesting · quant backtest · trading backtest · 策略回测 · 量化回测 · business automation · growth automation · AI business · 商业自动化 · 增长自动化 · AI商业

When to use / 何时使用

Use this skill when the user explicitly requests strategy backtest, or when the requested outcome clearly matches this workflow.

Do not activate it merely because the topic is related. Confirm the input, desired output, and any consequential external action.

Inputs / 输入

  • User-provided source material or connected authorized data
  • Desired output and constraints

Outputs / 输出

  • Primary requested artifact
  • Validation or review notes

Core stack / 核心工具

Market data / Python / backtesting library

Permissions / 权限

  • network:optional
  • filesystem:write

Treat permissions as capabilities to request or verify, not as permissions automatically granted by installing the skill.

Workflow / 工作流

  1. Inspect / 检查输入 — Verify source files/data, format, scope, and user constraints.
  2. Plan / 制定计划 — Select the deterministic tools, model calls, and external integrations needed.
  3. Preflight / 运行前检查 — Check dependencies, credentials, permissions, destination paths, and destructive-action risk.
  4. Execute / 执行 — Run the smallest reliable sequence of steps. Prefer deterministic code for calculations, parsing, rendering, and file operations.
  5. Validate / 验证 — Verify output structure, counts, schemas, file readability, test results, or other task-specific invariants.
  6. Review / 审核 — For subjective or consequential outputs, present a reviewable result before external publication, sending, payment, deployment, signing, or irreversible action.
  7. Report / 汇报 — State what was produced, where it was saved, what was verified, and what still needs human review.

Dependency suggestions / 依赖建议

  • No universal runtime package is required by the skill format itself. Install only the tools named in Core stack when your chosen implementation needs them.

The Agent Skill itself should remain installable even when optional runtimes are absent. Report missing runtime dependencies instead of silently installing privileged software.

Install / 安装

Inspect before installing

npx skills add BryceYuuu/open-workflow-skills --list

Install this skill into the current project

npx skills add BryceYuuu/open-workflow-skills --skill strategy-backtest

Install for a specific supported agent

npx skills add BryceYuuu/open-workflow-skills --skill strategy-backtest --agent codex

npx skills is a community ecosystem CLI, not an OpenAI-official installer. Review third-party skills and scripts before granting shell, filesystem, network, or credential access.

Example invocation / 调用示例

中文

使用 strategy-backtest 工作流处理我提供的输入。先检查依赖和权限,不覆盖源文件;执行后给我输出路径、验证结果和需要人工审核的部分。

English

Use the strategy-backtest workflow for the provided input. Check dependencies and permissions first, preserve the source, validate the result, and report outputs plus any human-review items.

Validation / 验证

At minimum:

  • Verify all expected outputs exist and are readable.
  • Verify the source was not unintentionally overwritten.
  • Verify structured outputs against their expected schema or invariants.
  • Record warnings, skipped steps, external dependencies, and failed checks.
  • Never mark the workflow successful solely because a model produced text.

See references/QUALITY.md for skill-specific quality expectations.

Safety / 安全

  • Do not invent source facts or claim an external action succeeded without verification.
  • Preserve source files unless the user explicitly requests in-place modification.
  • Surface missing dependencies, permissions, credentials, and failed checks clearly.
  • Treat outputs as analysis, not guaranteed financial outcomes; never execute live trades by default.
  • Label generated outputs as candidates or assisted results and require human review before consequential use.

Failure handling / 失败处理

If a dependency, credential, connected service, browser session, file, or required permission is unavailable:

  1. Stop the affected step.
  2. Preserve completed safe outputs.
  3. State exactly what is missing.
  4. Do not fabricate a successful result.
  5. Suggest the minimum next action needed to continue.

Files / 文件

  • SKILL.md — workflow instructions
  • references/QUALITY.md — quality/evaluation guidance
  • tests/cases.yaml — example evaluation cases

Implementation scripts may be added under scripts/ when deterministic execution materially improves reliability.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.