Skills Janitor vs Voyage AI

Side-by-side comparison of features, pricing, and ratings

Analysis reviewed Live tool data as of 2026-09-29
Cross-checked through our multi-step verification ·
Saved

At a glance

DimensionSkills JanitorVoyage AI
PricingFree, open-sourceContact sales (no free tier)
Primary UseCLI tool to manage/delete Claude Code/Codex skillsDomain-specific embedding & reranker models for enterprise RAG
Target UserClaude Code power users with many installed skillsEnterprises building RAG pipelines (finance, legal, code)
Key FeatureInteractive swipe-to-delete TUI, duplicate detection, token cost trackingLow-dim embeddings (3x-8x shorter), 32K context, domain fine-tuning
DeploymentLocal CLI, offline, no telemetryAPI-based, cloud or on-prem (contact)
ComplianceNone (local, no data collection)SOC 2, HIPAA

Choose Voyage AI if you're an enterprise building RAG pipelines with domain-specific retrieval needs (finance, legal) and have the budget. Choose Skills Janitor if you're a Claude Code user drowning in skills and want a free, interactive CLI tool to clean up context waste. They serve completely different purposes, so the decision is driven by your AI stack and budget.

Try Skills JanitorTry Voyage AI
Skills Janitor
Skills Janitor

Free MIT-licensed CLI that swipes away dead Claude Code and Codex skills and MCP servers to reclaim your context window.

Visit Website
Voyage AI
Voyage AI

Domain-tuned embedding models and rerankers from MongoDB for high-accuracy enterprise RAG retrieval.

Visit Website
Pricing
Free
Contact Sales
Plans
$0
—
Popularity
7 views
7.4k views
Skill Level
Intermediate
Intermediate
API Available
Platforms
CLI
WebAPI
Categories
💻 Code & Development
🗄️ Vector Databases & Retrieval
Features
Interactive Tinder-style swipe TUI for skills and MCP servers: left delete, right keep, down skip, u undo, i inspect
Swipe deck sorted most-likely-waste first, so heaviest-and-least-used skills and never-called MCP servers rank high
MCP server triage (v1.7): inventories user, project, .mcp.json, and plugin-bundled servers
Cross-references MCP servers against real usage from session transcripts (mcp__server__tool records)
Swiping an MCP server left removes its config entry with a timestamped .bak backup
Cross-scope duplicate detection using Jaccard similarity above a 30% threshold
Token cost reporting splitting always-loaded descriptions from on-demand skill bodies
Usage counts pulled from real session transcripts, including skills Claude auto-triggered
Health check for inventory, duplicates, and broken skills via /janitor-report, with --brief for inventory only
Auto-fix broken symlinks and empty directories with --prune via /janitor-fix
Prompt injection heuristic scan (v1.6+): instruction-override, hidden instructions, zero-width unicode, base64 payloads
Pre-install security and overlap scan via /janitor-discover before you install a skill
Natural language queries: 'check my skills', 'which skills are wasting context?', 'find an n8n skill'
Plugin skills and plugin-bundled MCP servers flagged for plugin review rather than individually deletable
Works with Claude Code and OpenAI Codex; zero dependencies beyond Bash, Python 3, and curl; offline-capable with no telemetry
General-purpose embedding models including voyage-3.5 and voyage-3.5 lite
Domain-specific embedding models optimized for finance, legal, and code
Company-specific fine-tuned embedding models on proprietary data
Voyage 4 model series for improved retrieval quality
voyage-multimodal-3.5 for multimodal retrieval across images and text
Low-dimensional embeddings (3x-8x shorter vectors) cut storage and search costs
32K-token context for long-document embedding
rerank-2.5 and rerank-2.5-lite with instruction following
voyage-context-3 for chunk-level detail with global document context
Batch API for large-scale embedding workloads
4x smaller model with faster inference
2x cheaper inference with superior accuracy
Modular design: plug-and-play with any vector DB and any LLM
SOC 2 and HIPAA compliance
Deployment via major clouds, SaaS customer tenants (in-VPC), and custom/on-premise

What real users say: Skills Janitor vs Voyage AI

Not marketing copy and not our opinion — a structured sweep of public discussion (reviews, forums, communities and video comments), showing what people praise and what they complain about for each tool.

Skills Janitor

19 mentions across 3 sources · 76% positive (averaged across 3 sources)

Product Hunt, GitHub, Lemmy

What users praise

  • • Interactive swipe interface makes cleanup fast and easy.
  • • Scans user, project, codex, and plugin skill locations comprehensively.
  • • Shows token cost per skill so you prioritize high-waste items.
  • • Sorts heaviest and least-used skills first for maximum impact.

What frustrates them

  • • Only supports Claude Code and OpenAI Codex — no other agents.
  • • No overlap percentage for similar skills, just dedup detection.
  • • Initial release missed /plugin install skills, requiring a fix.
  • • Usage stats rely on local log parsing; proxy setups may break it.

Researched Jul 3, 2026

Voyage AI

71 mentions across 6 sources · 38% positive — critical (weighted across 6 sources)

Hacker News, YouTube, App Store, Stack Overflow, GitHub, Lemmy

What users praise

  • • Domain-specific finance, legal, and code embedders beat general-purpose models on jargon-heavy corpora
  • • 3x-8x shorter embeddings cut vector storage and search costs without obvious accuracy loss
  • • 32K-token context handles long documents that force chunking in other models
  • • Rerank-2.5's instruction following lets you steer ranking behavior in plain language

What frustrates them

  • • Default terms grant Voyage a perpetual license to train on your API data
  • • No public pricing — everything routes through a sales conversation
  • • Not the fastest at scale; a Jina model reportedly beat it in one benchmark
  • • MongoDB ownership is steering the roadmap toward Atlas-first integration

Researched Sep 29, 2026

Who should pick which

  • Enterprise RAG engineer at a fintech firm
    Pick: Voyage AI

    Needs domain-specific embeddings for financial documents, long-context (32K tokens), and high-accuracy reranking. Voyage AI offers specialized models for finance and supports SOC 2/HIPAA compliance.

  • Claude Code power user with 100+ installed skills
    Pick: Skills Janitor

    Skills Janitor's interactive swipe TUI (janitor-swipe) lets you quickly delete the heaviest, least-used skills, cutting token waste by 30-40% for free.

  • Solo developer building a simple RAG app with free tools
    Pick: Skills Janitor

    Voyage AI has no free tier; Skills Janitor is free and immediately useful if you use Claude Code. But for RAG, Voyage AI's pricing may be prohibitive; consider alternatives.

  • ML team needing multimodal retrieval (images + text)
    Pick: Voyage AI

    Voyage AI announced voyage-multimodal-3.5, a dedicated multimodal model, which Skills Janitor does not offer.

  • Freelancer wanting to minimize Claude Code context costs
    Pick: Skills Janitor

    Skills Janitor's janitor-value tracks token costs per skill, helping you identify and remove expensive skills for free.

Frequently Asked Questions

Skills Janitor vs Voyage AI: which should you choose?

Choose Voyage AI if you're an enterprise building RAG pipelines with domain-specific retrieval needs (finance, legal) and have the budget. Choose Skills Janitor if you're a Claude Code user drowning in skills and want a free, interactive CLI tool to clean up context waste. They serve completely different purposes, so the decision is driven by your AI stack and budget.

Can I use Voyage AI for free?

No. Voyage AI requires contacting sales for pricing; there is no free tier or self-service plan.

Is Skills Janitor only for Claude Code?

It also supports OpenAI Codex (skills definitions). It scans user, project, codex, and plugin skill locations.

Does Voyage AI offer on-premise deployment?

Based on the facts, it's API-based cloud or on-prem (contact sales). The features mention 'SOC 2 and HIPAA compliance,' often requiring on-prem options.

Can Skills Janitor delete skills from all locations?

Yes. It deletes skills from user, project, codex, and plugin locations, except plugin skills can't be deleted individually (flagged).

Which tool supports long-context embeddings?

Voyage AI supports up to 32K tokens. Skills Janitor does not provide embeddings; it's a skill management CLI.

Are both tools open-source?

Skills Janitor is free and open-source. Voyage AI is proprietary, API-based with contact-based pricing.

Which tool is better for RAG pipelines?

Voyage AI, as it provides domain-specialized embedding models and rerankers designed for enterprise RAG. Skills Janitor has no relevance to RAG.

Does Skills Janitor work offline?

Yes. It is offline-capable and has no telemetry, according to the facts.

More Skills Janitor or Voyage AI comparisons

AI-Search vs Voyage AI comparison

Voyage AI and AI-Search serve completely different needs. Voyage AI is a specialized enterprise tool for high-accuracy embeddings and rerankers in RAG pipelines, ideal if you need domain-specific mode

Nubase vs Voyage AI comparison

Choose Voyage AI if you need domain-specific, high-accuracy embeddings and rerankers for enterprise RAG (finance, legal, code) with SOC 2/HIPAA compliance — expect sales-led pricing and modular integr

gitlab-duo-provisioning-blueprint vs Voyage AI comparison

Choose Voyage AI if your core need is high-accuracy retrieval on domain-specific data (finance, legal) with long-context support and low storage costs. Choose gitlab-duo-provisioning-blueprint if you

sandboxd vs Voyage AI comparison

If your need is high-accuracy retrieval over dense domain-specific documents (finance, legal, code), Voyage AI's specialized embedding models and rerankers are unmatched, but be prepared for enterpris

coreai-model-zoo vs Voyage AI comparison

These tools serve completely different needs. Choose Voyage AI if you run an enterprise RAG pipeline needing domain-tuned embeddings and rerankers, especially for finance/legal; its 32K context and lo

agentteam-email vs Voyage AI comparison

Voyage AI and agentteam-email solve completely different problems: Voyage AI is for high-accuracy retrieval in RAG (embedding/reranking), while agentteam-email manages email infrastructure for AI agen

Explore each tool further

Skills Janitor
View Skills Janitor reviewSkills Janitor alternatives
Voyage AI
View Voyage AI reviewVoyage AI alternatives

Browse these categories

Best AI Code & Development toolsBest AI Vector Databases & Retrieval tools
Still deciding? Get the weekly AI tools brief

One email a week — new tools, honest comparisons, no spam.

Last reviewed: July 3, 2026