CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

testland/experiment-sdk-testing

Use when code instrumented with Statsig, Optimizely, Split.io (Harness FME), Amplitude Experiment or VWO needs tests. Initializes the SDK offline from a committed config fixture, pins each test to an arm and checks assignment integrity, with per-vendor references.

Quality

92%

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

Overview
Quality
Evals
Security
Files

Quality

Content

92%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.

Excellent content: a lean, well-structured overview with a genuinely executable worked example, an explicitly sequenced workflow with built-in integrity checkpoints, and clean one-level-deep reference splitting across the five vendors. The only notable gap is that the inline example skips the teardown step the skill itself mandates.

Suggestions

Add the Step 5 teardown (e.g., client teardown in a pytest fixture or afterAll) to the worked example so it demonstrates the full five-step pattern it prescribes.

Show or link how to export the Optimizely datafile fixture, since Step 1 (commit the fixture) is asserted but not demonstrated.

DimensionReasoningScore

Conciseness

The body is lean throughout: a routing table replaces prose, no basic concepts (what an SDK or test is) are explained, and every section (shared-mechanism insight, five-step pattern, worked example, anti-patterns, limitations) adds information Claude does not already have, matching the 'every token earns its place' anchor.

5 / 5

Actionability

The Optimizely example is complete, executable Python and the routing table gives exact API names per vendor, but the inline example omits the skill's own Step 5 (teardown) and how to export the fixture — 'mostly executable guidance; concrete code with minor gaps' rather than fully covering the common case.

4 / 5

Workflow Clarity

Five explicitly numbered steps plus two mandated validation checkpoints ('Determinism' and 'Distribution' integrity tests 'every suite should carry') and an anti-patterns table mapping each failure mode to its fix, matching the anchor for clear sequence with explicit validation and checklists; the destructive/batch cap does not apply since test authorship is neither.

5 / 5

Progressive Disclosure

The SKILL.md body is a clear overview and all seven bundle files (five per-vendor references plus two supplementary recipe/example files) exist, are one level deep, and are each signaled with their purpose in the routing table and References section, matching the 'well-signaled one-level-deep references' anchor.

5 / 5

Total

19

/

20

Passed

Description

87%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.

A strong description with explicit 'Use when' triggers naming all five supported vendors and a concrete third-person statement of what the skill does. Keyword coverage is good but misses the generic natural terms ('feature flag', 'A/B test') a user might say without naming a vendor.

Suggestions

Add generic trigger synonyms such as 'feature flags' and 'A/B tests' so the description fires when users mention experimentation without naming a vendor.

Mention the determinism/distribution integrity checks or teardown in the capability list to close the minor specificity gap.

DimensionReasoningScore

Specificity

Names three concrete actions ('Initializes the SDK offline from a committed config fixture', 'pins each test to an arm', 'checks assignment integrity') but has minor coverage gaps (teardown, determinism/distribution checks are unmentioned), matching the 'several specific actions; minor gaps' anchor rather than the comprehensive-coverage anchor.

4 / 5

Completeness

Explicitly answers 'when' ('Use when code instrumented with Statsig... or VWO needs tests') with concrete vendor trigger phrases and 'what' with three specific third-person actions, matching the anchor for clearly answering both; it exceeds anchor 4 because the triggers are concrete, not merely explicit.

5 / 5

Trigger Term Quality

All five vendor names (Statsig, Optimizely, Split.io, Amplitude Experiment, VWO) plus 'needs tests' and 'SDK' are terms users would naturally say, but common synonyms like 'feature flag' and 'A/B test' are missing, so it fits 'good keyword coverage; a few natural terms missing' rather than comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

A clear niche — testing code instrumented with five named experimentation SDKs — with distinct vendor-name triggers and minimal overlap with generic testing skills, matching the 'clear niche with distinct triggers' anchor.

5 / 5

Total

18

/

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.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Reviewed

On this page