CtrlK
BlogDocsLog inGet started
Tessl Logo

sharaf/software-architecture-audit

Run an evidence-grounded software architecture audit workflow that builds a repo brief, selects single-auditor or specialist-panel mode, inspects boundary, layering, dependency, composition, cohesion, and testability risks, writes required finding blocks, and sequences incremental refactors. Use when asked for an architecture audit, architecture review, repo-structure review, software architecture report, audit_report.md, structural issue findings, or specialist-panel synthesis across multi-module systems.

100

1.85x
Quality

100%

Does it follow best practices?

Impact

100%

1.85x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Evaluation results

100%

91%

Architecture Audit: OrderFlow Order Management Platform

Correct report section structure and finding format

Criteria
Without context
With context

Section: Repo Brief

0%

100%

Section: Highest-Leverage Issues

0%

100%

Section: Boundary and Layering

0%

100%

Section: Dependency and Composition

0%

100%

Section: Testability and Change Friction

0%

100%

Section: Improvement Sequence

0%

100%

Section: Open Evidence Gaps

0%

100%

Section order

0%

100%

Highest-Leverage as summary

0%

100%

Finding block: Severity field

0%

100%

Finding block: Evidence field

0%

100%

Finding block: Why it matters

0%

100%

Finding block: Recommended improvement

0%

100%

Finding block: Likely refactor surface

0%

100%

Improvement Sequence format

0%

100%

No scorecards

50%

100%

No strengths section

100%

100%

100%

13%

Architecture Review: Ledger Service

Audit mode selection and repo brief completeness

Criteria
Without context
With context

Single-auditor mode

50%

100%

Repo brief: product name

100%

100%

Repo brief: language/framework

80%

100%

Repo brief: audit scope

100%

100%

Repo brief: docs checked

100%

100%

Repo brief: code surfaces checked

100%

100%

Repo brief: module map

100%

100%

Repo brief: test layout

100%

100%

Severity levels used

100%

100%

Evidence from actual files

100%

100%

Open Evidence Gaps section present

50%

100%

Ambiguous items in Open Evidence Gaps

80%

100%

Incremental recommendations

100%

100%

100%

39%

Architecture Review: Nexus CMS

Specialist-panel synthesis, deduplication, and forbidden recommendations

Criteria
Without context
With context

Specialist-panel mode

0%

100%

No scorecards

100%

100%

No strengths section

100%

100%

No heavyweight pattern prescription

100%

100%

Root cause deduplication

100%

100%

Findings in one section only

70%

100%

Evidence from actual files

100%

100%

Improvement Sequence: Goal field

0%

100%

Improvement Sequence: Why first field

0%

100%

Incremental first moves

100%

100%

Open Evidence Gaps present

0%

100%

Evaluated
Agent
Claude
Model
Claude Sonnet 4.6

Table of Contents