ai agent skill registries

Recent window: last 5.4 months (2026-05-01 → 2026-10-11), compared with the prior 5.4 months.

horizontal · 255 members · Data as of 2026-10-14

AI agent skill registries provide curated libraries, marketplaces, and management platforms for modular capabilities and specialized instructions that expand autonomous agents. They are designed for AI engineers, developers, and operators looking to equip digital workers and coding assistants with domain-specific abilities. Unlike general prompt directories or full agent builders, these registries focus specifically on packaging reusable, executable skill extensions for existing agent runtimes.

Metrics

Stage: crowded
Recent count: 226
Prior count: 21
Total count: 255
Momentum: 300.00
Momentum (normalized): 1.00
Attention: 0.36
Crowding: 0.99
Low crowding: 0.01
Concentration: 0.84
Low concentration: 0.16
Opportunity: 0.38

Source split

github: 190 (0.84)
hn: 14 (0.06)
ph: 21 (0.09)
yc: 1 (0.00)

Dominant source: github · Divergence: 0.84

Similar niches

modular skills for AI agents (0.89)
ai agent evaluation platforms (0.72)
ai agent development platforms (0.70)
multi-agent frameworks and protocols (0.65)
autonomous ai research agents (0.63)
reinforcement learning frameworks (0.62)
ai benchmarking platforms (0.61)
autonomous AI agent development tools (0.61)

Monthly trajectory

Pulled from real Hacker News comment threads on this niche's launches.

Most asked for

Add /s4h command for skill discovery (1)
Raised on skills-for-humanity
Add more PlusCal examples as reference material (1)
Raised on TLA+ Workbench skill for coding agents (compat. with Vercel skills CLI)
Add prefix to all skills (s4h- or sfh-) (1)
Raised on skills-for-humanity
Auto-update mechanism for packaged plugins via MCP (1)
Raised on Sx
Better built-in type safety and typing support (1)
Raised on 15 yrs of Django in prod: patterns I keep using (agent skills)
Better testing utilities for layer isolation (1)
Raised on 15 yrs of Django in prod: patterns I keep using (agent skills)
CI integration to pin skill versions and fail builds on eval score drops (1)
Raised on A package manager for agent skills with built-in evals
Clarify single vs. multiple skills for handling different event types (1)
Raised on Webhook Skills

Concerns raised

Agent skills as security vulnerability and LLM hallucination risk (1)
Raised on We build a Graph of public Skills
AI agents deployed in proprietary projects can't share fixes due to NDAs (1)
Raised on Shareful.ai
Another package manager may fragment ecosystem further (1)
Raised on Sx
Determining skill relevance and correlation (1)
Raised on We build a Graph of public Skills
Different event handling across products (SMS/Voice response differences) (1)
Raised on Webhook Skills
Different papers use different test sets and metrics, hard to compare (1)
Raised on Unlock the best engineering knowledge in papers for your coding agent
Difficulty finding skills without prefix (1)
Raised on skills-for-humanity
Difficulty tracking skill quality and drift across versions (1)
Raised on A package manager for agent skills with built-in evals

Compared to

Codex (2)
Raised on Shareful.ai, Distributing AI agent skills via NPM
aipack (similar tool on GitHub) (1)
Raised on Sx
Claude (1)
Raised on Distributing AI agent skills via NPM
CLI AI agents (1)
Raised on Shareful.ai
Cloudflare skills repo approach (1)
Raised on We build a Graph of public Skills
Existing package managers (npm, etc.) (1)
Raised on Sx
FastAPI (1)
Raised on 15 yrs of Django in prod: patterns I keep using (agent skills)
Gemini-cli (1)
Raised on Distributing AI agent skills via NPM

Members

Name	Source	Upvotes ▲	Launched
Perplexity Computer Skills	PH	456	2026-03-13
jev-skill	GITHUB	475	2026-09-20
reelbench-skills	GITHUB	695	2026-09-11
replica-skill	GITHUB	796	2026-10-03
yomiyasu	GITHUB	1371	2026-09-30