review-maintainability

Maintainability and design review instructions for the expert agent

Install with Tessl CLI

npx tessl i github:athal7/dotfiles --skill review-maintainability

What are skills?

1.45x

Quality

46%

Does it follow best practices?

Impact

99%

1.45x

Average score across 6 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./dot_config/opencode/skill/review-maintainability/SKILL.md

SKILL.md

Review

Evals

Evaluation results

100%

20%

Authentication Service PR Review

Phase 1 exploration before findings

Criteria

Without context

With context

Exploration log present

100%

Full file read logged

50%

100%

Caller search logged

53%

100%

AGENTS.md read logged

100%

Test file read logged

100%

JSON findings block

100%

JSON escalations block

100%

Debug print flagged

100%

Duplicate helper flagged

100%

Vague variable name flagged

100%

Without context: $0.2322 · 1m · 10 turns · 13 in / 3,881 out tokens

With context: $0.5741 · 2m 21s · 26 turns · 31 in / 8,394 out tokens

94%

Product Catalog Search PR Review

Issue context scope creep and test coverage

Criteria

Without context

With context

Scope creep detected

100%

Missing case-insensitive test

100%

Exploration log present

100%

Full file read logged

100%

Test file read logged

100%

JSON findings key present

100%

JSON escalations key present

100%

Scope creep finding references issue

100%

Findings in scope only

Without context: $0.1847 · 1m 4s · 7 turns · 12 in / 3,519 out tokens

With context: $0.4889 · 2m 23s · 22 turns · 59 in / 7,234 out tokens

100%

65%

Report Exporter PR Review

JSON output format escalations and finding tone

Criteria

Without context

With context

JSON object structure

100%

subprocess not in findings

100%

Correctness escalated

100%

I statements or questions

100%

Valid severity values

100%

suggested_fix on every finding

100%

Vague naming flagged

30%

100%

Duplicate logic flagged

100%

Missing tests flagged

100%

Without context: $0.2169 · 1m 26s · 10 turns · 15 in / 4,418 out tokens

With context: $0.5711 · 2m 32s · 23 turns · 61 in / 8,816 out tokens

100%

45%

Inventory Service PR Review

Pre-existing issue detection via git blame

Criteria

Without context

With context

git blame step logged

38%

100%

Pre-existing severity used

22%

100%

Old issues tagged pre-existing

53%

100%

Debug print flagged as new

100%

DRY violation flagged as new

100%

Exploration log present

100%

Missing export_csv test flagged

100%

Valid JSON structure

28%

100%

Without context: $0.3201 · 1m 37s · 15 turns · 20 in / 6,123 out tokens

With context: $0.6919 · 2m 38s · 34 turns · 286 in / 8,489 out tokens

100%

20%

Sessions Controller PR Review

Style tolerance and avoiding false positives

Criteria

Without context

With context

CONVENTIONS.md read logged

100%

handle_auth NOT flagged

100%

temp_token NOT flagged

100%

Missing test flagged

100%

Test file read logged

100%

Valid JSON output

100%

I statements or questions

100%

Without context: $0.3252 · 1m 36s · 16 turns · 21 in / 5,510 out tokens

With context: $0.5311 · 2m 38s · 19 turns · 22 in / 8,815 out tokens

100%

24%

Report Delivery Feature PR Review

Single responsibility, drift risk, and minimize diff

Criteria

Without context

With context

Single responsibility flagged

81%

100%

Whitespace diff flagged

100%

Drift risk flagged

100%

Exploration log present

100%

InvoiceBuilder read logged

100%

Security/correctness not in findings

100%

Finding titles 2-5 words

100%

I statements or questions

100%

Without context: $0.2294 · 1m 23s · 9 turns · 14 in / 4,322 out tokens

With context: $0.5397 · 2m 37s · 19 turns · 22 in / 9,455 out tokens

Evaluated: 1 day ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Authentication Service PR Review Product Catalog Search PR Review Report Exporter PR Review Inventory Service PR Review Sessions Controller PR Review Report Delivery Feature PR Review

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.