Patterns and techniques for evaluating and improving AI agent outputs. Use this skill when: - Implementing self-critique and reflection loops - Building evaluator-optimizer pipelines for quality-critical generation - Cre
Security
Low Risk
Quality
Good · 75
Install
ready
Audit version
audit-standard-v2
Source metrics come from upstream registries/repositories. Platform metrics come from user actions on SkillTrust and are protected by dedupe/rate-limit anti-abuse rules.
Patterns and techniques for evaluating and improving AI agent outputs. Use this skill when: - Implementing self-critique and reflection loops - Building evaluator-optimizer pipelines for quality-critical generation - Creating test-driven code refinement workflows - Designing rubric-based or LLM-as-judge evaluation systems - Adding iterative improvement to agent outputs (code, reports, analysis) - Measuring and improving agent response quality Imported from Smithery listing metadata.
• Add explicit When to Use / Guidelines sections.
• Provide at least one concrete input-output example.
• Publish versioned changelog and update cadence.
• Document compatibility and breaking-change policy.
• Expand capability limits and boundary conditions.
Is this a security certification?
No. SkillTrust audit is advisory and should be combined with your own review controls.
Can I install directly from this page?
No one-click install is provided. Use command guidance and run in your controlled environment.
Imported from Smithery listing metadata.
npx -y @smithery/cli install github/agentic-evalInstall method: registry-install
Quick install
Install paths
registry accessremote metadata pullruntime dependencies may be requiredPreflight checks
Post-install signal
Installed successfully? Send an activation signal to improve ranking quality over time.
Audit score 20 / 100. Risk guidance is advisory only; review evidence before install.
• Add troubleshooting and FAQ for common failures.
• Improve discoverability through verified source channels.
• Publish usage examples to increase activation quality.
• Reduce shell/file/network scope and document least privilege.
• Expose explicit provenance, dependency pinning, and security notes.
• Provide structured install + rollback steps for each supported agent.
• Mark official/verified status and keep metadata timestamps fresh.
Moderately fresh; review release notes before install.
Audit guidance: 20 / 100 · low
Scores with similar values can still differ in confidence; use evidence and risk lines below for final install judgment.
Risk 20 · Δ +0 · Findings 3
Apr 3, 2026 · auto
Latest low risk result from audit-standard-v2.