boundary-review

Audits codebases along subsystem boundaries and cross-boundary seams with scored review reports.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/fusuyfusuy/dot-agents --skill boundary-review-fusuyfusuy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: boundary-review
Source: https://github.com/fusuyfusuy/dot-agents/tree/main/skills/boundary-review
Command: npx skills add https://github.com/fusuyfusuy/dot-agents --skill boundary-review-fusuyfusuy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Deep architectural reviews often miss interface defects because auditors focus inside modules while contract drift hides at the seams between subsystems. This Skill decomposes a codebase into horizontal scopes plus a dedicated seam audit, producing standardized scores and a consolidated report without touching code until the operator approves. ## Core Features & Use Cases - Boundary-First Decomposition: Partitions a repository into 3-6 cohesive horizontal scopes plus one cross-boundary seam auditor covering API, persistence, event, config, and auth contracts. - Standardized Scoring Calibration: Applies a fixed severity rubric (Critical <7.0 through Exemplary 9.5-10) so scores stay consistent across agents and passes. - Mandatory Approval Gate: Writes per-scope markdown reports and a master audit report, then stops for explicit operator approval before any remediation with two-phase machine verification. - Use Case: Before a major release, run the protocol against a full-stack web app to catch drift between backend OpenAPI schemas and frontend typed clients, then approve only the P1 security fixes. ## Quick Start Ask the agent to run a boundary review of this repository, partition it into scopes, audit the seams, and present the scorecard for approval.

Frequently Asked Questions about boundary-review

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

FAQPage Schema
How do I run an architectural code review across subsystem boundaries?▼

Partition the codebase into 3-6 cohesive horizontal scopes, audit each scope internally, then run a dedicated seam audit across API, persistence, event, config, and auth contracts. The protocol synthesizes per-scope reports into a master audit report with a prioritized remediation roadmap.

What is a seam audit in code review?▼

A seam audit inspects only the contract interfaces between subsystems, such as backend route schemas versus frontend clients, database migrations versus ORM models, and event producers versus consumers. It targets cross-boundary drift that per-module reviews typically miss.

Does this review protocol work without mimori installed?▼

Yes, the protocol is harness- and tool-agnostic. Without mimori it falls back to rg, grep, ast-grep, directory trees, and targeted line-range reads for mapping, slicing, and call-graph traversal.

Can the skill modify code automatically after the audit?▼

No, audits are strictly diagnostic by default. The protocol stops at a mandatory approval gate after presenting the scorecard, and only executes remediations after explicit operator confirmation with two-phase machine verification.

Does it support single-agent environments without subagents?▼

Yes, single-agent harnesses execute the scopes sequentially, writing each scope's findings to .agents/reviews/ before moving on. Multi-agent harnesses can dispatch all auditors concurrently in one flat batch with degraded-scope fault tolerance.