dtsong/agentic-council/scout-library-evaluation

Library Evaluation

Purpose

Produce a structured, weighted comparison of candidate libraries to make dependency decisions based on evidence rather than familiarity or hype.

Scope Constraints

• Evaluates individual libraries and packages, not full frameworks or platforms.
• Focuses on technical fitness for the project, not market positioning or competitive landscape.
• Does not perform security audits; flag security concerns for handoff to the appropriate skill.

Inputs

• The need or problem the library should solve
• Any known candidates (or let the process discover them)
• Project constraints (bundle size budget, license requirements, framework compatibility)
• Current tech stack and existing dependencies

Input Sanitization

No user-provided values are used in commands or file paths. All inputs are treated as read-only analysis targets.

Procedure

Step 1: Identify Candidate Libraries

• Search npm, GitHub, and community recommendations for the problem space
• Include the most popular option, the most recently trending option, and at least one lightweight alternative
• Note any candidates already in use in the codebase's dependency tree

Step 2: Evaluate Popularity Metrics

For each candidate:

• npm weekly downloads (absolute number and trend direction)
• GitHub stars and recent star velocity
• Community size (Discord, Stack Overflow tags, GitHub Discussions activity)
• Notable adopters (large companies or well-known projects using it)

Step 3: Assess Maintenance Health

For each candidate:

• Commit frequency: Active development or maintenance-only?
• Issue response time: Median time to first response on new issues
• Release cadence: Regular releases or long gaps?
• Bus factor: Single maintainer or team? Corporate backing?
• Breaking change history: How are major versions handled?

Step 4: Measure Bundle Impact

For each candidate:

• Gzipped bundle size (via bundlephobia or similar)
• Tree-shaking support (ESM exports, side-effect-free)
• Dependency count (transitive dependencies and their health)
• Runtime performance characteristics if applicable

Step 5: Review API Quality and Developer Experience

For each candidate:

• TypeScript support (built-in types, DefinitelyTyped, quality of types)
• Documentation quality (getting started guide, API reference, examples)
• API ergonomics (intuitive naming, sensible defaults, composability)
• Migration path from current solution if applicable

Step 6: Check License Compatibility

For each candidate:

• License type (MIT, Apache 2.0, ISC, GPL, etc.)
• Compatibility with project's license
• Transitive dependency license concerns
• Any commercial use restrictions

Step 7: Produce Weighted Comparison Score

• Assign weights based on project priorities (e.g., bundle size matters more for client-side)
• Score each criterion 1-5
• Calculate weighted total
• Flag any deal-breakers that override the score

Progress Checklist

• [ ] Step 1: Candidates identified
• [ ] Step 2: Popularity metrics evaluated
• [ ] Step 3: Maintenance health assessed
• [ ] Step 4: Bundle impact measured
• [ ] Step 5: API quality reviewed
• [ ] Step 6: License compatibility checked
• [ ] Step 7: Weighted comparison scored

Compaction resilience: If context was compacted, re-read this SKILL.md and check the Progress Checklist for completed steps before continuing.

Handoff

• If license or security concerns emerge, recommend loading skeptic/threat-model for threat analysis.
• If the evaluation reveals broader framework or platform adoption questions, recommend loading scout/technology-radar.

Output Format

Library Comparison Matrix

Recommendation

Recommended: [Library Name] Rationale: [2-3 sentence justification referencing the top differentiating factors] Risks: [Known risks or caveats with the recommendation] Migration notes: [If replacing an existing dependency, key migration steps]

Quality Checks

• [ ] At least 3 candidate libraries evaluated
• [ ] npm download trends checked (not just absolute numbers)
• [ ] Maintenance health assessed (commit frequency, issue response, bus factor)
• [ ] Bundle size measured with gzipped numbers
• [ ] TypeScript support quality verified
• [ ] License compatibility confirmed
• [ ] Weighted scores reflect project-specific priorities
• [ ] Deal-breakers explicitly called out

Evolution Notes

<!-- Observations appended after each use -->