Audit prompts, CLAUDE.md, and agent SDK settings with five-axis PASS/WARN/FAIL scoring.
npx skills add https://github.com/TeamSPWK/nova-algorithm --skill claude-filter
Please help me install this Agent Skill. Skill: claude-filter Source: https://github.com/TeamSPWK/nova-algorithm/tree/main/skills/claude-filter Command: npx skills add https://github.com/TeamSPWK/nova-algorithm --skill claude-filter
Audits prompts, CLAUDE.md, and agent configurations to ensure they meet production-ready design principles, providing structured feedback and concrete improvements.
Paste your system prompts, CLAUDE.md, and agent settings to run a production-grade diagnostic.
High-intent search queries and answers about installing and using this skill.
Audit agent configurations for production readiness by applying a five-axis diagnostic model covering instruction structure, verification loops, context/cost, memory/state, and permission gating to generate explicit PASS/WARN/FAIL scores and structured improvement feedback.
A prompt audit for multi-agent orchestration evaluates system prompts and agent SDK settings across deployment scenarios to enforce safety requirements, validate reliability, and reduce deployment risk using structured toxicity checks and scoring.
Check CLAUDE.md files for LLM safety by running them through a diagnostic framework that tests instruction structure, verification loops, and permission gating, returning deterministic structured outputs to guide improvements.
Yes, the production prompt diagnostic enforces toxicity checks and delivers deterministic, structured outputs that can be consumed by humans or machines to guide concrete improvements for agent settings.
The best way to evaluate system prompts for deployment risk is applying a five-axis diagnostic model with explicit PASS/WARN/FAIL scoring to validate instruction structure, context/cost limits, and tool gating.
Run a prompt audit on agent settings before deployment to validate production-grade design principles, ensure multi-agent orchestration reliability, and enforce safety requirements across your configurations.
ClubMediterraneeGenerate Claude Code agent configurations from user requirements.
lucasmiachon-blipAudit medical slides for assertion-evidence, accessibility, data accuracy, and design-system conformance.
vasilyu1983Generate PRDs and validation templates for AI agent workflows with MCP integration.
JaiminVadadoriyaTest AI features including LLM prompts, tool calls, and nondeterministic outputs.
