Agent SkillSource: gstack / Garry Tan
Design Review
Loads one expert workflow for visual QA, hierarchy, states, and design polish.
designreviewagent skill
AportoSearchable skills and capabilities for AI agents
Find the Right AI Skill.Run the Right Capability.
Aporto gives agents a searchable marketplace of expert workflows and executable AI capabilities. Agents discover what they need at runtime, load only one skill when needed, and execute real work without dumping the catalog into context.
2 skills2 capabilitiesLoaded on demand, not in agent context
CapabilityRuns outside context
Screenshot QA
Capture a page, inspect layout issues, and return evidence your agent can use.
browserqaartifact
CapabilityNano Banana, Flux, Ideogram
Image Generation
Generate or edit visual assets without wiring separate model providers.
mediaimagerouting
Agent SkillSource: Aporto
LinkedIn Identity Research
Find or verify professional identity from email, name, company, or role signals.
researchlinkedinworkflow
Trending
90agent skill sources indexed today
1000+executable capabilities behind one router
0catalog pages loaded into the prompt by default
Connect your agent
One MCP connection. Pick your agent, paste one line, start searching.
Your agent sees a small Aporto tool surface. The marketplace stays server-side until the agent searches for a skill or capability.
Claude Code
Detailed setup guideAPORTO_API_KEY=sk-live-... codex mcp add aporto --url https://app.aporto.tech/api/mcp --bearer-token-env-var APORTO_API_KEYTwo-layer catalog
Skills teach the workflow. Capabilities do the work.
Aporto searches both layers. An agent can load a source-backed expert workflow, then run model, media, search, scraping, document, or automation capabilities through the same gateway.
Expert instructions with sources
Design Review by gstack / Garry Tan, LinkedIn Identity Research by Aporto, and future source variants from maintainers.
Executable AI actions
Image generation, video, TTS, web search, scraping, LLM routing, document parsing, commerce, and automation.
Capability marketplace
Give agents the ability to generate, search, scrape, transcribe, route models, and automate.
Aporto handles provider routing, async runs, and artifacts behind one interface.
180+ capabilities
Generate, edit, upscale, reframe, remove backgrounds, and route prompts across visual models.
Nano BananaNano Banana 2Nano Banana 2 ProFluxGPT ImagesIdeogramRecraftQwen ImageSeedreamTopaz Upscale
Terminal
$ npm install -g @aporto-tech/sdk $ export APORTO_API_KEY=sk-live-... $ aporto run "generate product image" \ --param prompt="a cat on the moon" \ --wait
Who is this for?
Aporto is built for teams turning AI agents into production workers.
Use it when your agent needs to do something outside the chat window: call models, create media, gather data, run jobs, and return usable outputs.
Agent builders who need real execution, not a prompt catalog
Open-source maintainers adding paid AI capabilities to repos
Startups replacing many AI APIs with one routed interface
Teams automating research, media, data, and operations
Creators who want production media workflows without provider setup
Smart Execution Engine
Every request is matched to an executable capability in real time.
- Discovers only the capabilities relevant to the task
- Runs the selected capability outside the model context
- Returns structured results, URLs, and artifacts
- Retries or falls back when an execution path fails
- Tracks cost, latency, and run history
Full control included
Aporto is not a black box.
You can define the routing behavior that fits each workflow.
- Pin allowed capabilities per workflow
- Define routing rules for agents and apps
- Control cost vs quality tradeoffs
- View execution logs
- Debug every request
- Keep backend provider details internal
Why Aporto instead of direct APIs
One integration replaces brittle provider-by-provider plumbing and giant tool lists.
Direct APIAporto
A giant prompt or tool catalogRuntime discovery only when needed
1000 MCP tools in the agentOne markdown skill and five MCP tools
Provider-by-provider keysOne Aporto API key
Hardcoded model workflowsDynamic capability routing
Manual polling and storage glueManaged runs, artifacts, and logs
FAQ
Answers for builders, creators, and teams.
What is Aporto in simple terms?
Aporto is an AI execution layer that connects your apps, workflows, or agents to 1000+ executable capabilities. Instead of manually integrating tools, you send one request and Aporto routes it to the best capability automatically.
Why do I need this if I can just use AI tools directly?
Single tools do not scale. Without Aporto, you integrate each tool separately, decide manually what to use, handle failures yourself, and maintain multiple systems. With Aporto, one integration replaces many tools, capabilities are selected automatically, failures are handled for you, and your system scales without extra infrastructure.
Do I lose control over what is happening?
No. You gain control. Aporto can route automatically by default, follow your exact rules, or run fully deterministic workflows. You can override decisions, pin specific capabilities, and inspect execution logs.
What if Aporto chooses the wrong capability?
You can prevent that with rules per task type, cost vs quality preferences, allowed capability groups, and fallback logic. Every execution is visible, traceable, and explainable.
Can I use Aporto without coding?
Yes. Non-developers can create content workflows, automate repetitive tasks, generate media and scripts, and connect AI tools together. Developers can go deeper and build full AI systems.
Will this increase my costs?
Usually no, and often the opposite. Aporto helps reduce cost by choosing cheaper capabilities when possible, avoiding unnecessary high-cost execution, reducing redundant integrations, and improving efficiency per request. You only pay for actual usage.
What makes Aporto different from other AI APIs?
Most APIs give access to a single capability. Aporto gives you a network of executable capabilities, automatic routing between them, the ability to combine them into workflows, and scalable execution infrastructure.
Pricing note
Free to start. Pay only when execution actually runs.
No setup feesNo long-term contractsScale as you grow
Aporto is not a prompt library. It is the execution layer between your agent and 1000+ AI capabilities.
Start Building