CtrlK
BlogDocsLog inGet started
Tessl Logo

evaluation

Build evaluation frameworks for agent systems. Use when testing agent performance systematically, validating context engineering choices, or measuring improvements over time.

53

Quality

60%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/AI-Agents-Safe-Coding-Skills-claude/skills/evaluation/SKILL.md

The canonical home for this skill is evaluation in administrakt0r/AI-Agents-Safe-Coding-Skills

SKILL.md
Quality
Evals
Security

Quality

Content

42%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 content covers the topic broadly with useful conceptual framing and a couple of illustrative examples, but is held back by verbosity, non-executable code, a workflow lacking validation checkpoints, and broken/vague references. It reads more as a primer than lean, actionable guidance.

Suggestions

Trim conceptual exposition Claude already knows (e.g., definitions of non-determinism, the duplicated '## References' block, and the dated metadata footer) and remove time-sensitive model names to improve conciseness.

Replace the stubbed evaluate_agent_response example with executable code that defines load_rubric/assess_dimension/weighted_average, or clearly mark it as illustrative pseudocode with a pointer to a real implementation.

Add validation/feedback-loop checkpoints to the 'Building Evaluation Frameworks' workflow (e.g., verify rubric scores before thresholding, re-run on regressions) and either create the referenced Metrics Reference file or remove the dangling reference.

DimensionReasoningScore

Conciseness

The body is noticeably verbose, spending many lines explaining concepts Claude already knows (non-determinism, multiple valid paths) and including padding such as a duplicated '## References' section, a dated metadata block, and time-sensitive model names, beyond just 'some' unnecessary explanation.

2 / 5

Actionability

There is some concrete guidance (rubric dimensions, a test-set structure example), but the code examples are not executable — load_rubric(), assess_dimension(), and weighted_average() are undefined stubs — leaving key implementation details missing.

3 / 5

Workflow Clarity

The 'Building Evaluation Frameworks' section gives a coherent 8-step sequence, but it has no validation checkpoints or feedback loops for a batch evaluation pipeline, which the rubric caps at 3.

3 / 5

Progressive Disclosure

Section structure is present and reasonable, but the referenced 'Metrics Reference' file does not exist, the '## References' section is duplicated and contains only vague bullet points with no links, and reference-style content is inlined.

3 / 5

Total

11

/

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 is strong: it clearly states both what the skill does and when to use it, with reasonably natural trigger phrases. The main weakness is specificity — only one concrete action verb is given, leaving the capability list less comprehensive than top examples.

DimensionReasoningScore

Specificity

The description names the domain ('Build evaluation frameworks for agent systems') with one concrete action verb plus three trigger gerunds, but does not list several specific actions, so it sits at the '1-2 concrete actions' anchor rather than comprehensive.

3 / 5

Completeness

It explicitly answers both 'what' ('Build evaluation frameworks for agent systems') and 'when' ('Use when testing... validating... measuring...') with concrete trigger phrases.

5 / 5

Trigger Term Quality

Trigger phrases like 'testing agent performance systematically', 'validating context engineering choices', and 'measuring improvements over time' are natural things a user would say, though common synonyms like 'evaluating agents' or 'agent benchmarks' are missing.

4 / 5

Distinctiveness Conflict Risk

'Evaluation frameworks for agent systems' is a fairly specific niche with mostly distinct triggers, though broad testing/benchmarking phrasing creates minor overlap with general testing skills.

4 / 5

Total

16

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
administrakt0r/AI-Agents-Safe-Coding-Skills
Reviewed

Table of Contents

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.