linear-propose

name	linear-propose
description	Analyze codebase to propose improvements, features, and fixes as Linear issues.

Linear Propose

Overview

Explore the project codebase to identify improvements, missing features, bugs, and tech debt. Present proposals to the user and create approved ones as Linear issues.

API Script

bun ~/.claude/skills/linear/scripts/linear-api.ts <command>

Process

1. Understand Scope

Check if the user specified a focus area (e.g., "performance", "UX", "security", a specific directory). If not, do a broad sweep.

2. Explore the Codebase

Systematically scan:

• Structure: package.json, directory layout, config files, routing
• Entry points: pages, routes, API endpoints
• Components & hooks: reusability, consistency, missing abstractions
• Error handling: missing try/catch, unhandled promise rejections, missing error boundaries
• Test coverage: untested critical paths, missing edge case tests
• Performance: heavy renders, missing memoization, N+1 queries, large bundles, unnecessary re-renders
• Security: exposed secrets, missing input validation, XSS/injection vectors, auth gaps
• Accessibility: missing ARIA attributes, keyboard navigation, screen reader support
• UX patterns: missing loading states, error states, empty states, optimistic updates
• Code quality: duplication, dead code, outdated dependencies, inconsistent patterns

Use Glob, Grep, and Read tools to explore. Be thorough but focused.

3. Cross-Reference Existing Issues

Fetch existing issues to avoid proposing duplicates:

source ~/.bashrc && bun ~/.claude/skills/linear/scripts/linear-api.ts list-issues --team <KEY> --limit 50
source ~/.bashrc && bun ~/.claude/skills/linear/scripts/linear-api.ts search-issues "<keyword>"

Skip proposals that overlap with existing issues.

4. Categorize & Prioritize Proposals

Group findings into categories:

Category	Label	Description
Bug	Bug	Something broken or behaving incorrectly
Performance	Performance	Optimization opportunities
Feature	Feature	Missing functionality the architecture suggests
Tech Debt	Tech Debt	Code quality, duplication, outdated patterns
Security	Security	Vulnerabilities or hardening opportunities
Accessibility	Accessibility	A11y improvements
UX	UX	User experience gaps

For each proposal, include:

### <Title> [<Category>]

**Priority**: 1-Urgent / 2-High / 3-Medium / 4-Low
**Effort**: S / M / L
**Files**: `path/to/file.ts`, `path/to/other.ts`

**Problem**: What's wrong or missing — be specific, reference code locations.
**Suggestion**: Recommended approach to fix/improve.

5. Return Proposals

Return all proposals as a structured list to the orchestrator. The orchestrator will present them and let the user pick which to create as Linear issues.

Key Principles

• Be specific: Reference exact file paths, function names, and line numbers
• Be actionable: Each proposal should be implementable as a single Linear issue
• Avoid noise: Don't propose trivial style nits or subjective preferences
• Respect existing patterns: Proposals should work within the project's conventions, not against them
• Deduplicate: Always check existing Linear issues before proposing