548 skills in the catalogue

New connections
for your agents.

Skills, tools and MCP servers — indexed from GitHub, sandbox-scanned and installable in one click into Cursor, Claude Desktop, Claude Code or any agent over HTTP.

Open the catalogue
Indexed skills
54829 publishers · 6 languages
Verified
0reviewed106 sandboxed
Installs
0via generated client config
Install targets
4Cursor · Claude Desktop · Claude Code · Any agent (HTTP)

How it works

From repository to agent in three steps.

A repository never reaches your agent directly: Synapth parses the manifest, runs it through the security scanner and only then emits a config your client understands.

Step 1

Crawl & parse

GitHub repositories with synapth.json, mcp-server.json or a SKILL.md are discovered by topic and filename, then normalised into one manifest.

> crawl(topic: 'mcp-server')
Step 2

Security scan

Every surface — prompt, tool schemas, README, entrypoint — runs through the static scanner. Risky patterns mean Sandbox; a clean pass means Community. Verified only comes from a human review.

[scan] prompt-injection · shell · secrets
Step 3

Client config

The manifest becomes the exact JSON for Cursor or Claude Desktop, a claude mcp add one-liner, or a minified context blob for agents calling the API.

mcpServers.<slug> = { command, args } ✓

Interactive request

Fire a real agent request

The button below calls the public gateway exactly the way an agent does and lights each node as the payload comes back.

Node 01Agent prompt"postgres"
Node 02Catalogue searchwaiting for a pulse
Node 03Sandbox scanbadge of the top hit
Node 04Agent contextsystem prompt + tool schemas
Request logStandby
> Gateway ready. Type a query and fire a pulse.
> Every line below is a real response from /api/v1/skills.

Trending in the catalogue

Browse all 548
thedotmack· 10h agoCommunity
wowerpoint

Prompts · TypeScript · v0.1.0

Turn one document into a kawaii NotebookLM slide-deck PDF. Use for "wowerpoint this", "make a deck about <file>", "turn this report into slides", or any request to render a single document as shareable narrative slides.

#ai#ai-agents#ai-memory

0 94.7K
thedotmack· 10h agoCommunity
what-the

Prompts · TypeScript · v0.1.0

What the? Use when the user wants a plain-English breakdown of something technical — the who, what, where, why, and when.

#ai#ai-agents#ai-memory

0 94.7K
thedotmack· 10h agoCommunity
weekly-digests

Prompts · TypeScript · v0.1.0

Generate a serial week-by-week narrative digest of a project's full claude-mem timeline. Splits the timeline into per-ISO-week files, then runs one consecutive subagent per week — each receiving the prior week's carry-forward block — to produce one chapter per ISO week of data. Use when asked for "weekly digests", "week-by-week story", "serial timeline", or "narrative chapters" of a project's history.

#ai#ai-agents#ai-memory

0 94.7K
thedotmack· 10h agoCommunity
version-bump

Prompts · TypeScript · v0.1.0

Automated semantic versioning and release workflow for Claude Code plugins. Handles version increments across package.json, marketplace.json, plugin.json manifests, build verification, git tagging, GitHub releases, and changelog generation. NPM publishing is the final human-required handoff because the maintainer raised npm security.

#ai#ai-agents#ai-memory

0 94.7K
thedotmack· 10h agoCommunity
timeline-report

Prompts · TypeScript · v0.1.0

Generate a "Journey Into [Project]" narrative report analyzing a project's entire development history from claude-mem's timeline. Use when asked for a timeline report, project history analysis, development journey, or full project report.

#ai#ai-agents#ai-memory

0 94.7K
thedotmack· 10h agoCommunity
standup

Prompts · TypeScript · v0.1.0

Facilitate a read-only standup across git worktrees, branches, or PRs to compare changes and produce one consolidation plan.

#ai#ai-agents#ai-memory

0 94.7K

What you get

One catalogue, every skill checked.

Built for agent stacks where a stray prompt injection or an unmetered tool loop is not an acceptable failure mode.

Catalogue

One catalogue

One index for MCP servers, prompts and HTTP tools. Full-text search with operators (category:MCP lang:python stars:>100), facets over language, author and tags, and a ranking that mixes install velocity, stars and retention.

MCP servers7
Prompts541
Tools0
MCP servers · Prompts · ToolsBrowse packages →
Security

Security scanner

Every manifest is scanned before it can be installed. The scanner never grants Verified on its own.

Rules:
prompt-injection · shell · secrets
Surfaces:
prompt · tools · readme · entrypoint
Sandbox:
106 flagged
Verified:
0 reviewed
Static analysisVerified = human review

Get started

Connect your agent to the catalogue.

Point your agent at the catalogue, or publish your own skill from a GitHub repository — parsed, scanned and listed in under a minute.

CursorClaude DesktopClaude CodeAny agent (HTTP)
bash
# Search the catalogue as an agent
$ curl -H 'X-Agent-Request: true' \
'https://synapth.localhost8081.ru/api/v1/skills?q=postgres'
# Fetch one skill with its manifest
$ curl https://synapth.localhost8081.ru/api/v1/skills/postgres-mcp
> 200 OK · manifest · install snippets