CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

vibe-ai-1/spec-as-source

Spec-as-source enforcement

23

Quality

28%

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

15%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 skill body is an empty stub — a TODO placeholder with no instructions, no workflow, and no structure. Every dimension that measures delivered guidance (actionability, workflow clarity, progressive disclosure) is at the floor; only conciseness avoids a minimum score, and only because there is nothing verbose about a one-line placeholder. The skill is non-functional as written.

Suggestions

Replace the TODO placeholder with a concrete, executable workflow for systematic debugging, e.g. "1. Reproduce the failure with a minimal input; 2. Form one hypothesis about the cause; 3. Instrument or bisect to test the hypothesis; 4. Apply the smallest fix; 5. Verify by re-running the originally failing case."

Add explicit validation checkpoints (re-run the failing test, confirm the fix, check for regressions) so the workflow includes a feedback loop rather than a linear guess-and-patch sequence.

Organize the body with section headers (e.g. ## Process, ## Common pitfalls) so the content is navigable; if detail grows beyond the quick-start level, split advanced material into one-level-deep reference files under references/.

DimensionReasoningScore

Conciseness

The body is a single placeholder line ("TODO: Add instructions for how the agent should execute this skill."), so there is no verbosity or padded explanation to penalize (ruling out anchors 1–2), but a TODO marker addressed to the skill author delivers no instructional value and no token earns its place, so it cannot score as lean, efficient content (anchors 4–5). It sits at the middle neutral point.

3 / 5

Actionability

The body contains zero concrete guidance — no code, commands, or steps — and is literally a placeholder acknowledging that instructions are absent, which is weaker than even anchor 1's example ("Use a PDF library to extract the text content...") that at least describes a task. Anchor 2 requires high-level hints, of which there are none.

1 / 5

Workflow Clarity

No steps exist in any form, matching anchor 1 ("Steps missing or incoherent; no sequence; no validation for risky operations"). Anchor 2 requires at least a rough sequence ("1. Open the document..."), which is absent — the body provides nothing to sequence.

1 / 5

Progressive Disclosure

The body has no sections, no structure, and no navigation, matching anchor 1's "no structure; impossible to navigate" clause; anchor 2 requires inlined content that belongs in separate files or buried references, and there is no content or references at all. No bundle files exist (no references/, scripts/, or assets/), and the simple-skill exception does not apply because a single TODO line is not "well-organized sections".

1 / 5

Total

6

/

20

Passed

Description

11%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 a single unmodified domain word that conveys no concrete capabilities, no usage triggers, and no scope boundaries. It is the weakest possible description short of being empty, scoring at or near the bottom of every dimension. The missing 'Use when...' clause also caps completeness regardless of any future rewording.

Suggestions

State concrete actions instead of the bare noun, e.g. "Guides systematic root-cause analysis of software bugs: reproduce the failure, form hypotheses, isolate the cause, and verify the fix."

Add an explicit 'Use when...' clause with natural trigger phrases users actually say: "Use when the user mentions debugging, a bug, a failing test, an error, a stack trace, a crash, or troubleshooting code."

Narrow the scope (e.g. to code debugging with a stated method) so the skill is distinguishable from testing, code-review, and general error-fixing skills that would currently all compete to trigger.

DimensionReasoningScore

Specificity

The description is the single word "Debugging" — a bare domain noun with no actions whatsoever, making it less informative than the anchor-1 example "Helps with documents" (which at least supplies a verb); it does not reach anchor 2 ("Processes PDF files"), which names a domain plus at least a generic action.

1 / 5

Completeness

The 'what' is an extremely vague single noun and the 'when' is entirely absent; anchor 1 ("missing both what and when, or both are extremely vague") fits, and the anchor-1 example "Helps with documents" is strictly more informative than "Debugging". It does not reach anchor 2's "vague 'what' and no 'when'" bar because even the 'what' conveys no relationship between the skill and any task.

1 / 5

Trigger Term Quality

"Debugging" is one natural keyword users do say, so anchor 1 ("no natural keywords") does not fit, but it supplies exactly one generic term and omits the natural phrases users actually use (debug, fix this bug, error, stack trace, crash, troubleshoot), matching anchor 2's "one or two generic keywords; missing the natural phrases users say".

2 / 5

Distinctiveness Conflict Risk

"Debugging" is scoped to one very broad domain, so it will not trigger for unrelated skills (not anchor 1's "conflict with virtually any skill"), but within that domain it is maximally generic with high overlap risk against testing, code-review, and error-fixing skills — exactly anchor 2's "very broad; high overlap risk with many similar skills" (cf. "Helps with document files").

2 / 5

Total

6

/

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

description_field

'description' is very short (9 chars), consider making it more detailed

Warning

Total

15

/

16

Passed

Reviewed

On this page