agentsclimarketplace

Eval

Skill iabhisekbosepm/claude-god-setup/.claude/skills/eval

Multi-agent orchestration system with 21 specialized agents, automated workflows, and cost-optimized model routing.

Install
npx -y skills add iabhisekbosepm/claude-god-setup --skill eval

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate a complete Langfuse evaluation pipeline: auto-detect app type, create datasets, LLM-as-judge evaluators, scoring configs, and CI integration.

SKILL.md

1.6 KB, as published. Nobody here has run it

/eval — Langfuse Evaluation Pipeline Generator

Generate evaluation setup for: $ARGUMENTS

  1. Discover — agents/eval-architect.md Step 1: Explore codebase, detect LLM app type (RAG, chatbot, agent, code-gen, summarization, classification, extraction), identify prompts, traces, and data flows
  2. Design — agents/eval-architect.md Steps 2-3: Select 4-8 eval dimensions from the dimension matrix, generate seed dataset with 10-20 items covering happy path, edge cases, and adversarial inputs
  3. Generate — agents/eval-architect.md Steps 4-5: Write complete evals/ directory: config.yaml, evaluators (LLM judge + heuristic), runner.py, ci_eval.py, .env.example
  4. Setup — agents/eval-architect.md Step 6: If LANGFUSE_SECRET_KEY is set, run setup_langfuse.py to create dataset and score configs in Langfuse. If not, skip gracefully with instructions.
  5. Document — agents/eval-architect.md Step 7: Generate PM/SME-friendly README with quick start, architecture, and troubleshooting
  6. Verify — Run verification checklist: correct app type? dimensions appropriate? dataset validates against schema? scripts importable? README clear?

Rules: Always include at least one heuristic evaluator. Never hardcode API keys. README must be understandable by non-developers.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.