review-correctness-and-data-integrity

Review a change for whether it does what it is meant to do, and whether the data it touches survives it intact. Use as one lens in a code review run.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./code-review/skills/review-correctness-and-data-integrity/SKILL.md

The canonical home for this skill is tessl/code-review

Quality

Content

86%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is an excellent, lean review lens skill that provides highly specific, actionable guidance for correctness and data integrity review without wasting tokens on concepts Claude already understands. The four-section structure (Scope, Method, Threshold, Reporting) creates a natural workflow. The only meaningful improvement would be a brief concrete example of a finding formatted per the Reporting section's requirements.

Dimension	Reasoning	Score
Conciseness	Every sentence carries unique, non-obvious guidance. No explanation of what code review is, no padding, no concepts Claude already knows. The content is dense with domain-specific heuristics (e.g., 'application code can be redeployed but written data may require repair') that genuinely add value.	5 / 5
Actionability	The guidance is concrete and specific—trace write paths, check interrupted runs, read the AGENTS.md chain, name the boundary and breaking change. However, it's instruction-only with no worked examples showing what a finding looks like in practice, which would elevate it further. Still, for an instruction-only review lens skill, the specificity is strong.	4 / 5
Workflow Clarity	The skill follows a clear implicit sequence: Scope (what to look at) → Method (how to analyze) → Threshold (what to report) → Reporting (how to report). The method section sequences three distinct analysis passes. Minor gap: no explicit validation checkpoint or feedback loop, but this is a review lens (read-only analysis), not a destructive operation, so the cap at 3 for missing validation does not apply.	4 / 5
Progressive Disclosure	This is a focused, single-purpose skill under 50 lines with clean section organization (Scope, Method, Threshold, Reporting). No external references are needed and none are artificially created. The sections are well-signaled and logically ordered for easy navigation.	5 / 5
	Total	18 / 20 Passed

Description

37%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 conveys a general sense of purpose—reviewing code changes for correctness and data integrity—but lacks concrete actions, specific trigger terms, and explicit 'when to use' guidance. The literary phrasing ('data it touches survives it intact') sacrifices clarity for style, and the description would benefit significantly from concrete keywords and explicit trigger scenarios.

Suggestions

Add explicit trigger terms users would naturally use, such as 'pull request', 'PR review', 'diff', 'data integrity', 'correctness', 'regression check'.

Replace vague phrasing like 'does what it is meant to do' and 'data survives intact' with concrete actions such as 'verifies functional correctness against requirements, checks for data corruption or loss in migrations/transformations'.

Add a clear 'Use when...' clause with specific scenarios, e.g., 'Use when reviewing PRs or diffs that involve data transformations, schema changes, or business logic modifications'.

Dimension	Reasoning	Score
Specificity	The description names the domain (code review) and mentions two concerns ('does what it is meant to do' and 'data survives intact'), but these are vague and abstract rather than concrete actions. No specific techniques, tools, or outputs are mentioned.	2 / 5
Completeness	The 'what' is present but vague ('review a change for correctness and data integrity'). The 'when' is weakly implied with 'Use as one lens in a code review run' but lacks explicit trigger phrases or concrete scenarios. The 'when' clause is too abstract to serve as effective routing guidance.	3 / 5
Trigger Term Quality	Contains 'code review' and 'change' as relevant keywords, but misses natural terms users would say like 'pull request', 'PR', 'diff', 'review my code', 'data integrity', 'correctness check', or 'regression'. The phrasing is indirect and literary rather than keyword-rich.	2 / 5
Distinctiveness Conflict Risk	The focus on correctness and data integrity during code review provides some differentiation, and the phrase 'one lens in a code review run' suggests it's part of a multi-skill review pipeline. However, the description is vague enough that it could overlap with general code review skills or testing-related skills.	3 / 5
	Total	10 / 20 Passed

Validation

100%

Warnings & errors only

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

Validation — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: tesslio/product-plugins
Path: code-review/skills/review-correctness-and-data-integrity/SKILL.md
Commit: 37f0a21

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

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.