Completion Evaluator

🌐Community
by masanao-ohba · vlatest · Repository

This tool assesses generated text completions for quality and relevance, streamlining workflows by automatically identifying strong outputs.

Install on your platform

We auto-selected Claude Code based on this skill’s supported platforms.

1

Run in terminal (recommended)

terminal
claude mcp add completion-evaluator npx -- -y @trustedskills/completion-evaluator
2

Or manually add to ~/.claude/settings.json

~/.claude/settings.json
{
  "mcpServers": {
    "completion-evaluator": {
      "command": "npx",
      "args": [
        "-y",
        "@trustedskills/completion-evaluator"
      ]
    }
  }
}

Requires Claude Code (claude CLI). Run claude --version to verify your install.

About This Skill

What it does

The completion-evaluator skill assesses the quality of text completions generated by AI models. It provides evaluations based on criteria such as relevance, coherence, and factual accuracy. This allows for iterative improvement of prompts and fine-tuning of language model behavior.

When to use it

  • Prompt refinement: Evaluate different prompt variations to identify which produces the best results from a language model.
  • Model comparison: Compare the output quality of multiple AI models given the same task.
  • Content generation review: Automatically assess generated content for accuracy and relevance before publication or further use.
  • Iterative development: Track improvements in completion quality as you adjust prompts or fine-tune a model.

Key capabilities

  • Evaluates text completions based on multiple criteria.
  • Provides scores reflecting the quality of each completion.
  • Supports comparison of different completions.

Example prompts

  • "Evaluate the following completion: [completion text]"
  • "Compare these two completions and tell me which is better: [completion 1] | [completion 2]"
  • "Assess this generated paragraph for factual accuracy: [paragraph text]"

Tips & gotchas

The effectiveness of the evaluation depends on clearly defining the desired qualities in your prompts. Consider providing specific examples or guidelines to guide the evaluation process.

Tags

🛡️

TrustedSkills Verification

Unlike other registries that point to live repositories, TrustedSkills pins every skill to a verified commit hash. This protects you from malicious updates — what you install today is exactly what was reviewed and verified.

Security Audits

Gen Agent Trust HubPass
SocketPass
SnykPass

Details

Version
vlatest
License
Author
masanao-ohba
Installs
5

🌐 Community

Passed automated security scans.