comprehension-gate

Reviews code diffs and PRs for system-level comprehension risks before merging.

Updated Jul 16, 2026
One-click install
npx skills add https://github.com/Cloud-Byte-Consulting/plugins --skill comprehension-gate-cloud-byte-consulting
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: comprehension-gate
Source: https://github.com/Cloud-Byte-Consulting/plugins/tree/main/prompt-workflows/skills/comprehension-gate
Command: npx skills add https://github.com/Cloud-Byte-Consulting/plugins --skill comprehension-gate-cloud-byte-consulting

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI-generated and human-written code changes often pass syntax and test checks while hiding system-level risks like credential exposure, cross-service side effects, and unbounded blast radius. This Skill acts as a senior-engineer reviewer that tests whether anyone actually understands what a change does before it ships. ## Core Features & Use Cases - Seven-Dimension Comprehension Analysis: Examines credential exposure, cross-service side effects, blast radius, state persistence, token/session TTLs, implicit assumptions, and overall comprehensibility. - Structured Comprehension Artifact: Produces a change summary, severity-ordered findings table, blast radius map, pre-merge questions, and a CLEAR / REVIEW REQUIRED / HOLD verdict. - Use Case: Before merging an AI-generated PR that touches a shared Redis cache, paste the diff and receive a findings table flagging that line 47 writes an auth token readable by the analytics service, plus questions that must be answered before shipping. ## Quick Start Paste a diff or PR description and ask the agent to run a comprehension gate review on it before merging.

Frequently Asked Questions about comprehension-gate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review an AI-generated PR before merging?

Paste the diff or PR description into the comprehension gate workflow, optionally adding system context like a module manifest or dependency description. It returns a findings table ordered by severity, a blast radius map, and questions that must be answered before merging.

What does a comprehension gate check in a code review?

It checks seven dimensions: credential and secret exposure, cross-service side effects, blast radius, state and persistence changes, token and session TTLs, implicit assumptions, and whether the change is understandable in isolation. It explicitly does not review style, naming, or test coverage.

Can I use this review without providing the actual diff?

Yes, a description of the change works, but the review will note where the actual code would enable more specific analysis. Providing the real diff plus system context such as dependency maps produces the most precise blast radius assessment.

When should I not rely on this kind of code review?

Do not use it as a substitute for linters, CI test suites, or style checks, since it deliberately ignores syntax and coverage. It also cannot assess blast radius without system context, so it will explicitly state when dependency information is missing.

What do the CLEAR, REVIEW REQUIRED, and HOLD verdicts mean?

CLEAR means intent matches behavior and blast radius is bounded. REVIEW REQUIRED means specific listed questions need answers before shipping. HOLD means the change shows signs the author may not understand system-level implications and needs senior engineer review of Critical findings.