Paths: File paths (shared/,references/,../ln-*) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root. Ifshared/is missing, fetch files via WebFetch fromhttps://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}.
ln-832-bundle-optimizer
Type: L3 Worker Category: 8XX Optimization
Reduces JavaScript or TypeScript bundle size using keep/discard verification. JS/TS projects only.
Overview
| Aspect | Details |
|---|---|
| Input | JS/TS project path plus optional optimization scope |
| Output | Smaller bundle plus a machine-readable modernization summary |
| Scope | JS/TS only |
Workflow
Phases: Pre-flight -> Baseline -> Analyze -> Optimize Loop -> Report
Phase 0: Pre-flight Checks
| Check | Required | Action if Missing |
|---|---|---|
package.json exists | Yes | Block optimization |
| Build command available | Yes | Block optimization |
Output directory (dist/ or build/) detectable | Yes | Build once to establish baseline |
| Workspace baseline safe | Yes | In managed runs coordinator already prepared it; in standalone runs protect rollback locally |
MANDATORY READ: Load shared/references/ci_tool_detection.md for build detection. MANDATORY READ: Load shared/references/mcp_tool_preferences.md and shared/references/mcp_integration_patterns.md
Use hex-graph first when hotspots or clone groups materially improve optimization targeting. Use hex-line first for local code and config reads when available. If MCP is unavailable, unsupported, or not indexed, continue with built-in Read/Grep/Glob/Bash and state the fallback in the summary.
Runtime Coordination
Managed runs receive deterministic runId and exact summaryArtifactPath from ln-830. Standalone runs remain supported; if runtime arguments are omitted, generate a standalone run-scoped artifact before returning.
Phase 1: Establish Baseline
1. Run the detected build command.
2. Measure total size of the output directory.
3. Record baseline bytes and per-chunk sizes.| Metric | How |
|---|---|
| Total size | Sum all bundle output files except source maps |
| Per-chunk sizes | Measure individual JS and CSS files |
| Source map handling | Exclude .map files |
Phase 2: Analyze Opportunities
| Check | Tool | What It Finds |
|---|---|---|
| Unused dependencies | depcheck or equivalent | Packages not imported anywhere |
| Bundle composition | Bundle analyzer | Large dependencies or duplicates |
| Tree-shaking gaps | Manual scan | Namespace imports or heavy entrypoints |
| Code splitting gaps | Route analysis | Lazy-load candidates |
Optimization categories:
| Category | Example | Typical Savings |
|---|---|---|
| Remove unused deps | dead dependencies | 10-50KB per package |
| Named imports | lodash-es named imports | 50-200KB |
| Lighter alternatives | replace heavy date or utility libs | 50-300KB |
| Dynamic imports | lazy-load heavy components | lower initial bundle |
| CSS optimization | remove unused CSS, minify | 10-100KB |
Phase 3: Optimize Loop (Keep/Discard)
Per-optimization cycle:
FOR each optimization:
1. APPLY the change
2. BUILD
IF build fails -> DISCARD and revert
3. MEASURE new bundle size
4. KEEP only if the bundle is smaller than the current baseline
5. LOG the resultStop conditions:
| Condition | Action |
|---|---|
| All optimizations processed | Stop and report |
| 3 consecutive discards | Stop for plateau |
| Build infrastructure breaks | Revert to last keep and stop |
| Bundle already below target | Stop and report |
Optimization order:
- remove unused dependencies
- improve tree-shaking and imports
- replace heavy libraries
- introduce code splitting
- optimize CSS
Phase 4: Report Results
| Field | Description |
|---|---|
project | Project path |
baseline_bytes | Original bundle size |
final_bytes | Final bundle size |
reduction_bytes | Bytes saved |
reduction_percent | Percentage reduction |
optimizations_applied | Count of kept optimizations |
optimizations_discarded | Count plus reasons |
deps_removed[] | Dependencies removed |
details[] | Per-optimization savings |
artifact_path | Durable worker report path, if written |
Configuration
Options:
build_command: "" # Auto-detect
output_dir: "" # Auto-detect
run_depcheck: true
run_bundle_analyzer: false
remove_unused_deps: true
fix_tree_shaking: true
replace_heavy_libraries: true
enable_code_splitting: trueError Handling
| Error | Cause | Solution |
|---|---|---|
| Build fails after optimization | Invalid code or config change | Revert that optimization |
| No size reduction | Optimization ineffective | Discard it |
| Analyzer unavailable | Tooling missing | Continue with manual inspection |
References
- optimization_categories.md
shared/references/ci_tool_detection.md
Runtime Summary Artifact
MANDATORY READ: Load shared/references/coordinator_summary_contract.md
Emit a modernization-worker summary envelope.
Managed mode:
ln-830passes deterministicrunIdand exactsummaryArtifactPath- write the summary to the provided
summaryArtifactPath
Standalone mode:
- omit
runIdandsummaryArtifactPath - write
.hex-skills/runtime-artifacts/runs/{run_id}/modernization-worker/ln-832--{identifier}.json
Monitor (2.1.98+): For build commands in keep/discard loop expected >30s, use Monitor. Fallback: Bash(run_in_background=true).
Definition of Done
- Baseline bundle size established from real build output
- Optimization opportunities analyzed from bundle composition and project usage
- Each optimization evaluated with keep/discard verification
- Only size-reducing optimizations kept
- Final report captures bundle delta, discarded attempts, and removed dependencies
modernization-workersummary artifact written to the managed or standalone path
Version: 1.0.0 Last Updated: 2026-03-08