← Back to Skills Marketplace
Progressive Validator
by
tltby12341
· GitHub ↗
· v1.0.1
· MIT-0
162
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install progressive-validator
Description
Multi-stage backtest validation framework — fail fast with short windows (smoke/stress/medium/full) before committing to expensive full-period backtests, sav...
Usage Guidance
This skill appears coherent and self-contained: it orchestrates backtest stages and stores local results. Before using it, (1) inspect validator.py yourself (it is included) and confirm you are comfortable with it writing validation_results.json in your working directory, (2) only run the suggested backtest submission command if your backtest-poller CLI is the legitimate tool you installed (don’t run a sibling ../backtest-poller/cli.py from an untrusted location), and (3) run the tool in an isolated directory or under version control to avoid accidental overwrite of files. If you want extra assurance, run the script in a disposable environment or review the backtest-poller code it calls.
Capability Analysis
Type: OpenClaw Skill
Name: progressive-validator
Version: 1.0.1
The progressive-validator skill is a legitimate framework for managing multi-stage backtest validation of trading strategies. The Python script (validator.py) manages state using a local JSON file and suggests shell commands for a separate backtest-poller utility; it does not perform any network requests, access sensitive system files, or execute arbitrary code. The instructions in SKILL.md are consistent with the tool's purpose and do not contain any prompt-injection attacks or malicious directives.
Capability Assessment
Purpose & Capability
Name/description match the included code and SKILL.md. The validator implements stage orchestration, local result persistence, and command suggestion for an external backtest CLI (backtest-poller). Required binary (python3) is appropriate and proportionate.
Instruction Scope
SKILL.md and validator.py confine actions to suggesting commands, recording stage results, and printing status. They do not read unrelated system files or environment variables. However, the runtime instructions assume you will execute an external script (../backtest-poller/cli.py); executing that script is outside this skill and could run arbitrary code if the external CLI is untrusted.
Install Mechanism
No install spec or remote downloads. The package is delivered as code files (validator.py, config example) and a requirements.txt declaring no dependencies. Nothing is written to disk by an installer beyond the user's explicit run of the script.
Credentials
The skill requests no environment variables or credentials and the code does not access secrets. The only external dependency is the backtest-poller CLI (path is user-specified); ensure that CLI is the legitimate tool you expect before running suggested commands.
Persistence & Privilege
The skill persists validation_results.json (configurable) in the current working directory and will overwrite if present. It does not request persistent platform privileges nor set always:true. Running in an isolated/workspace directory is recommended to avoid accidental file overwrite.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install progressive-validator - After installation, invoke the skill by name or use
/progressive-validator - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Clarified that this skill coordinates validation, but actual backtest submission is handled by the separate backtest-poller skill.
- Added concrete command examples for both submitting backtests (with backtest-poller) and recording results using validator.py.
- Updated documentation to specify the dependency on the backtest-poller skill for submission and early-stop functionality.
- Improved guidance on workflow: check "what to run next" and explicitly record each stage’s outcome.
- No changes to validation logic or core features. Documentation and interface instructions enhanced for ease of use.
v1.0.0
Initial release of progressive-validator — a multi-stage backtest validation framework:
- Introduces a four-stage validation pipeline (Smoke, Stress, Medium, Full) to catch bad strategies quickly and save compute time.
- Documents default time windows, drawdown thresholds, and recommended usage for each validation stage.
- Provides guidance on configuring backtest windows and skip rules based on the type of strategy change.
- Describes integration with early-stop logic to halt failing backtests automatically.
- Offers concrete usage instructions and sample scripts for running each stage.
- Lists time savings and validation rules to maximize efficiency and ensure process transparency.
Metadata
Frequently Asked Questions
What is Progressive Validator?
Multi-stage backtest validation framework — fail fast with short windows (smoke/stress/medium/full) before committing to expensive full-period backtests, sav... It is an AI Agent Skill for Claude Code / OpenClaw, with 162 downloads so far.
How do I install Progressive Validator?
Run "/install progressive-validator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Progressive Validator free?
Yes, Progressive Validator is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Progressive Validator support?
Progressive Validator is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Progressive Validator?
It is built and maintained by tltby12341 (@tltby12341); the current version is v1.0.1.
More Skills