CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

check

Independently check application acceptance criteria in a live browser or saved recording with the Vibium CLI. Use for a formal verification step in the development loop, with PASS, FAIL, or REVIEW and recorded evidence.

Quality

85%

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

SKILL.md
Quality
Evals
Security

Quality

Content

88%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, highly actionable operational guide: exact commands, flags, exit codes, and a result contract, wrapped in a well-validated six-step workflow with explicit feedback loops. Its only weaknesses are mild — a reference-like provider-config section inlined in the main file and a small amount of repetition.

DimensionReasoningScore

Conciseness

The body is dense and operational throughout (e.g. "Confirm `vibium check --help` works", "Leave `VIBIUM_AI_REASONING_EFFORT` unset for Anthropic/Google") with no explanation of concepts Claude already knows. Minor trimmable material keeps it at 4: the five-provider environment-variable matrix is reference-like, and "use the `browser` skill if installed" is repeated in two sections.

4 / 5

Actionability

Commands are copy-paste ready and cover the common cases: `vibium check --json "<claim>"`, `vibium check -i record.zip "<claim>"`, `vibium ready ai --json`, `-o verification.zip`, `--report verdict.json`, `--expect pass`, exact env vars, a worked example claim, and a complete JSON result contract with exit-code semantics. This matches the fully executable anchor.

5 / 5

Workflow Clarity

The numbered "Development loop" is clearly sequenced with explicit validation (read the JSON result and report the actual verdict), a feedback loop ("investigate the evidence, fix the cause, and rerun the original claim... Do not weaken the claim"), error-recovery guidance for REVIEW and each `reason` code, and an explicit stop condition. Setup is gated by `ready ai` exit-code semantics.

5 / 5

Progressive Disclosure

Sections are well organized (Setup, Development loop, Browser cleanup, Recording and report flags, Result contract) and the broader CLI reference is properly deferred ("use the `browser` skill if installed... this skill can run on its own using `vibium <command> --help"); no bundle files exist to verify. It is not 5 because the provider-configuration matrix is reference-like content inlined in the main file, and the under-50-line simple-skill exception does not apply to this ~130-line body.

4 / 5

Total

18

/

20

Passed

Description

75%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 solid description that clearly states what the skill does, names its tool and outputs, and includes an explicit "Use for" clause. Its main gaps are the absence of common trigger synonyms ("verify", "test") and an abstract rather than concrete when-clause.

Suggestions

Add concrete trigger phrases to the when-clause, e.g. "Use when the user asks to verify the app works, confirm a feature, or check acceptance criteria before finishing a task."

Include natural synonyms such as "verify", "test", or "confirm" alongside "check" so the description matches more of the phrasings users actually say.

Consider mentioning the saved-evidence use case in user terms (e.g. "when the user wants to re-verify an earlier recorded run") to sharpen distinctiveness from live browser-automation skills.

DimensionReasoningScore

Specificity

"Independently check application acceptance criteria in a live browser or saved recording with the Vibium CLI" plus "PASS, FAIL, or REVIEW and recorded evidence" names the tool, both verification modes, and concrete outputs. It falls short of 5 because it describes one core action with variants rather than a comprehensive list of multiple distinct capabilities.

4 / 5

Completeness

The what is clear (independent acceptance verification with verdicts and recorded evidence) and an explicit "Use for a formal verification step in the development loop" clause provides the when. Not 5 because the when-clause is abstract and lacks concrete user-side trigger phrases; the explicit "Use for" wording keeps it above the 3 cap.

4 / 5

Trigger Term Quality

Natural terms like "check", "verification", "acceptance criteria", "browser", and "recording" match what users would say. It is not 5 because common synonyms and variations such as "verify", "test", or "QA" are missing, but coverage is good rather than merely partial.

4 / 5

Distinctiveness Conflict Risk

Naming the "Vibium CLI" and framing the niche as independent acceptance verification distinguishes it from building and browser-automation skills. Minor overlap remains with generic testing/verification skills that also mention checking behavior in a live browser, so it is not 5.

4 / 5

Total

16

/

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.

Repository

VibiumDev/vibium

Path

skills/check/SKILL.md

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.