agent-evaluation

Evaluate agent behavior with versioned cases and explicit verifiers. Use when comparing agent or prompt changes, reproducing failures, or running agent regression tests.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/agent-evaluation/SKILL.md

Quality

Content

73%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body delivers a well-sequenced, validation-rich evaluation procedure with a real, well-linked reference bundle and executable supporting code. Its main weakness is conciseness: a time-sensitive editorial note and a verbose limitations section add tokens that do not help Claude act.

Suggestions

Move the 'Modified by AAS maintainers on 2026-09-05...' line out of the body (or into a changelog/deprecated section) — time-sensitive editorial notes penalize conciseness.

Tighten the Limitations list by merging related bullets (e.g. the retry and exception points) into fewer, denser statements.

Consider relocating the worked-uncertainty JS function into the architecture-sketches bundle so the SKILL.md body stays a lean overview, keeping only the interval interpretation and expected checks inline.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and does not explain basic concepts, but the modification-history line ('Modified by AAS maintainers on 2026-09-05...') is time-sensitive padding outside a deprecated section, and the limitations list could be tightened, so it sits at 'mostly efficient but could be tightened'.	3 / 5
Actionability	A concrete 6-step procedure, a worked example with specific cases, an executable Wilson-interval function, and an expected-output template give mostly executable guidance; some steps remain principle-level ('do not retry until green') rather than command-level, leaving minor gaps.	4 / 5
Workflow Clarity	The six steps are explicitly sequenced with validation checkpoints (step 2 'Validate the harness' with known-pass/known-fail/failure cases, 'fix the harness and repeat') and a fix-verify-rerun feedback loop (step 6), matching the anchor for clear sequence plus explicit validation and error-recovery loops.	5 / 5
Progressive Disclosure	The core procedure is inline and detailed architecture patterns are deferred to a real one-level-deep bundle (references/architecture-sketches.md) with five clearly-signaled, verified section anchors; the inline worked-uncertainty JS block is content that could arguably live in the bundle, leaving a minor organization gap.	4 / 5
	Total	16 / 20 Passed

Description

78%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description cleanly answers both what and when with concrete trigger phrases and a distinct niche. It is slightly thin on the breadth of concrete actions, which keeps specificity and distinctiveness just below the top anchors.

Dimension	Reasoning	Score
Specificity	It names the domain ('agent behavior') and two concrete instruments ('versioned cases', 'explicit verifiers') but does not list several specific actions, matching the '1-2 concrete actions, not comprehensive' anchor rather than the several-action anchor at 4.	3 / 5
Completeness	It explicitly states what ('Evaluate agent behavior with versioned cases and explicit verifiers') and when ('Use when comparing agent or prompt changes, reproducing failures, or running agent regression tests') with concrete trigger phrases, matching the top anchor.	5 / 5
Trigger Term Quality	'comparing agent or prompt changes', 'reproducing failures', 'agent regression tests' are natural phrases a user would say; good coverage, though a few synonyms (e.g. 'eval', 'A/B testing') are absent, so it stops short of the comprehensive anchor at 5.	4 / 5
Distinctiveness Conflict Risk	'versioned cases', 'explicit verifiers', and 'agent regression tests' carve a clear niche with low conflict risk; minor overlap remains with general testing skills, so it is just below the minimal-conflict anchor at 5.	4 / 5
	Total	16 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: sickn33/agentic-awesome-skills
Path: skills/agent-evaluation/SKILL.md
Commit: b1aebac

Reviewed: ~18 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.