v4.1.0 — Gemini App, Gemini CLI & Antigravity Support

Skillary v4.1.0

• Added support for Google Gemini App (Skills replacing Gems via SKILL.md and .skill bundle import)
• Added first-class support for Gemini CLI (~/.gemini/skills/) and Google Antigravity (.agents/skills/)
• CLI: Added --agent flag with automatic multi-agent path detection
• Expanded supported agents to 9 runtimes (Claude Code, Gemini App, Gemini CLI, Antigravity, Codex, Cursor, Cline, Roo Code, Copilot)
• Synchronized version 4.1.0 across Hub, skillary-agents, and all 12 category repositories

v3.0.0 — Focused Agent Skills & Multi-Agent Orchestration

I built Skillary because dumping hundreds of unvetted instructions into an agent's context breaks real workflows. When an agent has too much generic prompt text, it hallucinates boundaries, burns tokens, and loses focus on the task.

v3.0.0 restructures everything into modular, single-responsibility plugins and adds native multi-agent orchestration:

1. 277 Curated Skills Across 12 Domain Plugins

• Modular by design: Take only what you need. Zero kitchen-sink repositories.
• Strict linting & boundaries: Hard 200-character description limits, 30-character slug caps, mutual sibling disambiguation, and zero import-failure stubs.
• Audited against AI tells: Direct, plain technical instructions without generic filler.

2. Multi-Agent Orchestration with skillary-agents

Frontier models now have real subagent execution capabilities (Claude Code Task, Antigravity invoke_subagent, Cursor/Codex worktrees). They're no longer just autocomplete engines; they can coordinate, build, and verify autonomously.

• Host-neutral orchestrator protocol: A shared contract that works across Claude Code, Google Antigravity, and Codex CLI.
• 6 Zero-slug role templates: scout, planner, worker, verifier, security-auditor, and shipper. These define behavioral contracts and never hardcode skill names, so they survive tool renames.
• 8 Production playbooks: End-to-end task workflows for shipping features, codebase rescue, security reviews, and business launches.

3. Fast Local Loading & SQLite FTS5 Search

• Install any category or skill directly:

  npx skills add poorvith-mp/skills-developer
  npx skills add poorvith-mp/skills-developer/systematic-debugging

• Local SQLite FTS5 search index with zero runtime telemetry.

v2.1.0 — I said production-grade. 30 of them were error messages.

Back in August I tagged v2.0 and called it "315 production-grade skills." Then I actually read all 315 files.

30 of them weren't skills. Their entire body was Full .md body could not be fetched — an error message from a failed import, published as a product. If you installed chief-of-staff or blockchain-security-auditor, that's what you got.

That's the headline, but the quieter problems were worse because you'd never notice them.

What was actually broken

281 skills told Claude to check that their code compiles. Every finance skill. Every education skill. tax-strategist had a checklist item about unhandled promise rejections. One template got stamped across every domain and nobody read the result.

228 descriptions were too long to work. Claude cuts the description at ~250 characters when it's deciding whether to load a skill. Mine averaged 274, and the trigger clause was always the last sentence. So the only part that decides whether a skill fires was the part getting cut off.

180 of those triggers were generated filler. Use when the user asks about tax strategist — nobody types that. People describe a task, not a skill name.

The router pointed at 37 skills that don't exist and missed 210 that do. Its own docs say a stale index makes it "actively harmful." It was right.

What changed

All 30 stubs are written. Real process, real deliverables. The regulated ones (medical billing, loan files, privacy, ESG) say which jurisdiction they need and that a licensed human signs off.

All 315 descriptions rewritten, capability first, triggers inside 250 chars. Skills that overlap now point at each other, so financial-analyst says "Not for budget-vs-actual variance work, use fp-and-a-analyst" and vice versa. Before, whichever fired first was luck.

The checklists match their domain now. A finance skill checks that figures reconcile and assumptions are stated. An education skill checks that you didn't assess something you never taught.

It's installable now

That's the part I should've done first.

/plugin marketplace add prvthmpcypher/skillary
/plugin install skills-finance@skillary

Or for Codex, Cursor, Gemini CLI and about 70 other agents:

npx skills add prvthmpcypher/skills-finance

12 categories, install only what you want. The hub is renamed from open-claude-skills to skillary. Old URL redirects.

Why it won't rot again

Everything that used to be hand-maintained is generated: the 315-row index, the router's index, all 13 manifests, all 315 bundles. scripts/validate.py reports zero findings and CI runs it on every push, so a description over 250 chars fails the build.

I tested that by breaking things on purpose. Over-long description fails. Added skill fails the index check. Edited skill fails the bundle check.

CI also caught something I couldn't. Bundles built on my Windows machine embedded CRLF, and I was byte-comparing compressed zips across Python versions. Both passed locally and failed the second they hit Linux.

Still not done

No skill ships scripts/ or assets/ yet. All 315 are instructions, not executable capability. That's the next thing.

One skill from the old gap list, cost-finops-engineer, is still unbuilt.

---

315 skills, 12 categories, MIT. If a skill doesn't fire when it should, that's a bug now, open an issue.

v2.0: Major Audit & Modernization — 315 Production-Grade Agentic Skills

🚀 open-claude-skills v2.0

A major milestone release delivering a comprehensive audit, modernization, and taxonomy overhaul across the entire 13-repository ecosystem.

Highlights:

• 315 Production-Grade Skills: Streamlined from 342 legacy skills into 315 fully structured, procedural skills across 12 domain categories.
• 100% Agent Skills Spec Compliance: Every skill conforms to agentskills.io standards with YAML frontmatter triggers, step-by-step phased execution, quality verification checklists, and negative anti-pattern boundaries.
• 14 New 2026 AI Workflow Skills: Added high-demand capabilities addressing critical industry gaps (Legacy Code Modernization, AI Eval Suite Builder, Observability Engineering, Data Pipelines, AI Governance, FinOps, Video Scripting, and Knowledge Architecture).
• Taxonomy Rebalancing: Cleaned folder naming conventions, removed persona wrappers, merged overlapping skills, and relocated misplaced workflows into appropriate repositories.
• Progressive Disclosure Reference Files: Added dedicated domain reference guides for WCAG 2.2, OWASP Top 10, STRIDE, MITRE ATT&CK, OpenAPI 3.1, PCI-DSS v4.0, HIPAA, GDPR, and ISO 27001.

Category Repositories:

• skills-developer (80 skills)
• skills-marketing (50 skills)
• skills-specialized (37 skills)
• skills-design (29 skills)
• skills-business (27 skills)
• skills-gamedev (22 skills)
• skills-sales-support (20 skills)
• skills-education (16 skills)
• skills-finance (15 skills)
• skills-personal (11 skills)
• skills-writing (5 skills)
• skills-meta (3 skills)

Fixed bugs and added more skills

Fixed

• Total skill count corrected: 319 → 342
• Per-category counts corrected: Developer 82→88, Business 26→29, Marketing 57→60,
  Specialized 53→55, Design 26→29, Game Dev 20→22, Sales & Support 11→13, Finance 7→9
  (Education, Personal, Writing were already accurate)
• Added 29 previously-undocumented or newly-added skills to the full skill index
  across 11 category tables
• skills-meta now links to its actual repo in the category table (was unlinked)

Added

• CHANGELOG.md and CODE_OF_CONDUCT.md (were missing from this repo)

Initial Release of all Skills

311 Claude skills split into 11 separate repos by category so you can grab only what you need.