CtrlK
BlogDocsLog inGet started
Tessl Logo

analyze-comparison-tests

Collects comparison test run artifacts and answers the user's questions based on the trajectories of each run. WHEN TO USE: collect comparison test artifacts

52

Quality

57%

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 ./.github/skills/analyze-comparison-tests/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

68%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 is a lean, actionable two-step workflow with an executable command, an input specification, and a well-signaled reference to the report template. Its main weakness is the absence of validation/error-recovery for a batch download operation, which the rubric caps at 3 for workflow clarity.

Suggestions

Add a validation checkpoint after artifact collection: verify `comparison-artifacts/` is non-empty and contains the expected branch/stimulus/model subdirectories before proceeding to insight extraction.

Add an error-recovery branch (e.g. what to do if the collect script fails or the input JSON path is missing), since this is a batch operation that fails silently without verification.

Show a concrete example of how to read a trajectory markdown file and turn it into an answer, so the 'Extract insights' step is fully executable rather than implied.

DimensionReasoningScore

Conciseness

The body is mostly efficient: a tight two-step procedure with a concrete command and a structural diagram, with only minor over-explanation (e.g. spelling out what a trajectory file records).

4 / 5

Actionability

It provides an executable command (`npm run compare:collect -- input.json`), the expected input shape, output directory, and a folder-structure diagram, with only minor gaps (no example of how to actually read/parse a trajectory file).

4 / 5

Workflow Clarity

Two steps are clearly sequenced and there is a conditional check (skip collect if the directory exists), but this is a batch download operation with no validation that artifacts were retrieved successfully and no error-recovery checkpoint, which caps the score at 3 per the rubric.

3 / 5

Progressive Disclosure

Structure is clean: the body is an overview with the report template correctly delegated to a one-level-deep, clearly signaled reference (`./references/report-template.md`); the only minor gap is that the referenced template is quite thin.

4 / 5

Total

15

/

20

Passed

Description

46%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 states a clear purpose but its trigger clause is a near-verbatim restatement of the action rather than natural user-facing language, and it lacks concrete triggering conditions. It is reasonably distinct from other skills but reads more like a label than a discoverable prompt.

Suggestions

Rewrite the 'WHEN TO USE' clause into natural phrases a user would actually say, e.g. 'Use when the user asks to compare test runs, analyze comparison test artifacts, or wants a report across comparison runs.'

Replace the duplicated phrase ('collect comparison test artifacts') with synonyms and concrete triggers like 'comparison test reports', 'compare:run artifacts', or 'run trajectories'.

Add concrete trigger conditions (what questions, what comparison scenarios) so Claude can reliably decide when this skill applies.

DimensionReasoningScore

Specificity

It names the domain ('comparison test run artifacts') and two concrete actions ('collects ... artifacts', 'answers the user's questions based on the trajectories'), but the actions are generic rather than a comprehensive list of capabilities.

3 / 5

Completeness

It states a clear 'what' (collect artifacts and answer questions from trajectories), but the 'when' is just a bare 'WHEN TO USE: collect comparison test artifacts' that restates the action rather than giving explicit triggering conditions.

3 / 5

Trigger Term Quality

The trigger text 'collect comparison test artifacts' is a near-duplicate of the action phrase rather than natural user language, and there are no synonyms or common variations a user would actually say.

2 / 5

Distinctiveness Conflict Risk

'Comparison test run artifacts' and 'trajectories of each run' carve a fairly specific niche tied to a particular test workflow, with only minor overlap risk against general artifact-collection skills.

4 / 5

Total

12

/

20

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
microsoft/GitHub-Copilot-for-Azure
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.