homingo scan
Instant fleet health check. Runs locally in seconds with zero API calls — uses heuristic analysis to detect skill overlaps and scope overload.
Usage
homingo scan [options]What It Does
- Loads all skills from your configured
skillsDir - Computes pairwise overlap scores using keyword and name similarity (no LLM calls)
- Checks each skill for scope overload using heuristic pattern detection
- Produces a health score (0–100) and severity-ranked report
Why Use Scan?
scan is the fastest way to get signal from Homingo. While audit and lint use LLM-powered routing simulation for precise accuracy measurements, scan gives you an instant overview using local heuristics only.
Use scan when you want to:
- Get a quick health check after
homingo init— see results in under a second - Run in CI without API keys — scan needs no LLM provider
- Triage before a full audit — identify which pairs are worth testing with the Shadow Router
Output
- Terminal summary with color-coded overlap pairs and overload warnings
- Self-contained HTML report that auto-opens in your browser
The report includes:
- Fleet health score (0–100)
- Conflicting pairs ranked by overlap severity
- Scope overload findings with reasons
- Next steps pointing to
auditandlintfor deeper analysis
Health Score
The health score is a weighted composite:
| Component | Weight | What it measures |
|---|---|---|
| Conflict ratio | 70% | Fraction of total pairs flagged as CRITICAL or HIGH overlap |
| Overload ratio | 30% | Fraction of skills flagged as scope-overloaded |
A score of 100 means no critical overlaps and no overloaded skills. A score below 50 suggests significant fleet issues that need attention.
Overlap Severity
Pairs are classified by their heuristic overlap score:
| Severity | Overlap Score | Meaning |
|---|---|---|
| CRITICAL | ≥ 50% | Very likely to cause misrouting |
| HIGH | 35–49% | Probable routing confusion |
| MEDIUM | 20–34% | Worth monitoring |
| LOW | < 20% | Minimal overlap |
Flags
| Flag | Description | Default |
|---|---|---|
--skills-dir <path> | Path to skills directory | From config |
--all-pairs | Analyze every skill pair | Heuristic selection |
--enhanced | Enable bigram matching for better overlap detection | Off |
--json | Output JSON instead of terminal summary | Off |
--no-open | Don't auto-open the HTML report | Off |
Examples
# Quick scan with defaults
homingo scan
# Scan all pairs for a thorough overview
homingo scan --all-pairs
# Scan with bigram matching for improved accuracy
homingo scan --enhanced
# JSON output for CI pipelines
homingo scan --json --no-open
# Scan a custom skills directory
homingo scan --skills-dir ./my-skillsExit Codes
| Code | Meaning |
|---|---|
0 | Fleet is healthy — no critical overlaps or overloaded skills |
1 | Issues found — at least one CRITICAL/HIGH pair or overloaded skill |
TIP
Use exit code 1 in CI to catch fleet issues early — even without API keys. For precise routing accuracy measurements, follow up with homingo audit.
Scan vs Audit vs Lint
scan | audit | lint | |
|---|---|---|---|
| Speed | Instant (< 1s) | Minutes | Minutes |
| API calls | None | Many | Many |
| Accuracy | Heuristic overlap | LLM routing simulation | LLM routing simulation |
| Rewrites | No | No | Yes (--fix) |
| Best for | Quick triage, CI gates | Deep diagnostics | Pre-deploy validation |
