Skills evaluation
📦

evaluation

Content revision r2 Safe ⚙️ External commands🌐 Network access

Build Reliable Agent Evaluations

Agent quality is difficult to measure because runs vary and valid solutions follow different paths. This skill builds practical rubrics, test sets, and evaluation pipelines.

Supports: Claude Codex Code(CC)
🥉 78 Bronze

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "evaluation" from https://skillstore.io/skills/sickn33-evaluation.md and its manifest at https://skillstore.io/api/skills/sickn33-evaluation/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/sickn33-evaluation.md Signed manifest GET /api/skills/sickn33-evaluation/manifest Signed lockfile GET /api/skills/sickn33-evaluation/lockfile

Test it

Using "evaluation". Create a rubric for a research agent that returns cited answers.

Expected outcome:

  • Factual accuracy: 35 percent, scored from unsupported to fully verified.
  • Citation accuracy: 30 percent, based on claim and source agreement.
  • Completeness: 20 percent, based on required coverage.
  • Source quality: 10 percent, favoring relevant primary sources.
  • Tool efficiency: 5 percent, with a passing overall score of 0.80.

Using "evaluation". Design regression coverage for a customer support agent.

Expected outcome:

  • Simple cases test direct policy questions with one clear answer.
  • Medium cases test account context and multiple policy conditions.
  • Complex cases test ambiguity, escalation, and conflicting customer details.
  • Very complex cases test long conversations, context retention, and safe handoff.

Using "evaluation". Compare two context strategies for an agent.

Expected outcome:

Run both strategies on the same stratified test set. Compare weighted quality, token usage, tool calls, failure patterns, and confidence intervals.

Security Audit

Safe
v5 • 7/23/2026 Open versioned report

All nine static findings are false positives caused by Markdown fences, descriptive prose, integration names, and a source metadata URL. The skill is documentation-only and contains no executable commands, network requests, reconnaissance behavior, or prompt injection.

1
Files scanned
244
Lines analyzed
0
Review items
0
False positives ignored

Risk Factors

⚙️ External commands (3)
SKILL.md:150-158 SKILL.md:158-164 SKILL.md:164-192
🌐 Network access (1)
SKILL.md:5
No confirmed security findings were detected by the latest completed static and semantic audit. This does not prove the skill has no side effects.
Audited by: codex View Audit History →
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/sickn33-evaluation/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/sickn33-evaluation/security.svg)](https://skillstore.io/skills/sickn33-evaluation?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/sickn33-evaluation?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-evaluation/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/sickn33-evaluation.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA · BibTeX · CFF)

APA citation

sickn33. (2026). evaluation security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-evaluation/audits/5

BibTeX citation

@techreport{sickn33-sickn33-evaluation-2026, author = {sickn33}, title = {evaluation security audit report (audit version 5)}, institution = {Skillstore}, year = {2026}, number = {5}, url = {https://skillstore.io/skills/sickn33-evaluation/audits/5}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "evaluation security audit report (audit version 5)" version: "unspecified" type: report authors: - name: "sickn33" date-released: "2026-07-23" url: "https://skillstore.io/skills/sickn33-evaluation/audits/5" identifiers: - type: other value: "skillstore:sickn33-evaluation:audit:5" description: "Skillstore immutable audit report identifier"

Compare variants

4 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score
muratcankoylan Recommended

muratcankoylan-evaluation

Skillstore Score 81
Evidence Confidence High
Skillstore usage 11
Updated

2026-08-21

ChakshuGautam

chakshugautam-evaluation

Skillstore Score 80
Evidence Confidence High
Skillstore usage 9
Updated

2026-08-21

sickn33 Current

sickn33-evaluation

Skillstore Score 78
Evidence Confidence High
Skillstore usage 11
Updated

2026-08-21

Asmayaseen

asmayaseen-evaluation

Skillstore Score 77
Evidence Confidence High
Skillstore usage 12
Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: High
55
Architecture
85
Maintainability
87
Content
69
Community
83
Spec Compliance

What You Can Build

Validate a New Agent

Create a representative test set and weighted quality rubric before release.

Compare Context Strategies

Measure quality, token cost, and tool efficiency across competing context configurations.

Monitor Production Quality

Define sampled evaluations, regression alerts, and human review checkpoints for deployed agents.

Try These Prompts

Create a Basic Rubric 
Create an evaluation rubric for [agent task]. Include accuracy, completeness, and efficiency with clear scoring levels and a passing threshold.
Design a Test Set 
Design a test set for [agent system]. Cover simple through very complex tasks, realistic usage, known edge cases, and expected outcomes.
Compare Agent Configurations 
Build an experiment comparing [configuration A] and [configuration B]. Define controlled variables, quality metrics, token metrics, sample size, and decision criteria.
Design Continuous Evaluation 
Design a continuous evaluation pipeline for [production agent]. Include sampling, automated judging, human review, baselines, alerts, privacy controls, and regression reporting.

Best Practices

  • Evaluate end outcomes while allowing multiple valid execution paths.
  • Establish a baseline and use realistic test cases before comparing changes.
  • Combine automated scoring with systematic human review for edge cases.

Avoid

  • Do not judge quality with one metric or one successful example.
  • Do not overfit tests to a specific sequence of agent actions.
  • Do not deploy automated judges without checking bias and agreement with humans.

Frequently Asked Questions

Does this skill run evaluations automatically?
No. It provides methods and design guidance for building an evaluation system.
Can it evaluate non-deterministic agents?
Yes. It emphasizes outcomes, quality dimensions, and repeated representative tests instead of exact action sequences.
Which quality dimensions should I use?
Start with accuracy, completeness, citations, source quality, and tool efficiency. Adjust dimensions and weights for your task.
How large should the first test set be?
Begin with a small representative sample, then expand with real usage patterns, complexity levels, and known failures.
Can an LLM replace human evaluation?
No. LLM judges scale routine assessment, while human reviewers should validate edge cases, bias, and subtle failures.
How should I detect regressions?
Record baseline metrics, rerun the same stratified tests after changes, and investigate significant quality or efficiency declines.

Developer Details

Author

sickn33

License

MIT

Skillstore revision

r2

Version notice

The author did not declare a version.

Repository

https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/evaluation

Ref

88a8e9a07f4c54ab105c1c41b6267c287146b07b

Maintenance freshness

7/26/2026

Usage

9 downloads · 147 views

File structure

📄 SKILL.md

More from sickn33

View all

Related Skills

View all