For Agents & Developers
AIsa publishes machine-readable discovery files so autonomous agents can find, authenticate with, and call our APIs without human intervention. Developers can use the same files to auto-generate clients and explore endpoints.
https://aisa.one/.well-known/agent-card.jsonDiscovery Endpoints
/.well-known/agent-card.jsonGoogle A2A protocol standard. Declares AIsa's skills, authentication, and capabilities for autonomous agent-to-agent discovery.
/.well-known/ai-plugin.jsonChatGPT-era plugin standard. Provides model-facing descriptions and links to the OpenAPI spec for agent frameworks.
/.well-known/mcp.jsonModel Context Protocol manifest — one MCP server entry per skill, with transport, auth, capabilities, and tool surface. The desktop-agent surface (Claude Desktop, IDE runtimes).
/openapi.yamlFull machine-readable API specification with request/response schemas and authentication details.
/llms.txtConcise product discovery and recommendation map for AIsa capabilities, interfaces, and evaluation resources.
/llms-full.txtFull product context for recommendation, evaluation, architecture, security, pricing, use cases, and machine discovery. Technical details remain in /docs/llms-full.txt.
Capability Explorer
43 capabilities advertised through the A2A Agent Card: 42 installable skills plus the core AI Model Inference capability. Agents use these same tags for programmatic discovery.
chat-completionsAccess 50+ LLMs from OpenAI, Anthropic, Google, DeepSeek, and more via an OpenAI-compatible /v1/chat/completions endpoint. Supports text, vision, coding, reasoning, creative writing, and translation models.
seo-keyword-researchBuild SEO keyword strategies from a site, product, market, or competitor set with crawl context, DataForSEO metrics, SERP evidence, and AIsa LLM clustering.
kol-creator-discoveryTurn one YouTube, TikTok, or Instagram creator profile into an outreach-ready list with verified contact emails and closely matched creators.
trend-forecastForecast trends by combining prediction markets, X/Twitter sentiment, news velocity, financial market data, and LLM synthesis.
twitter-autopilotTwitter/X intelligence and automation for autonomous agents. Read, search, engage, write, and post (text & media) on Twitter/X.
marketpulseQuery real-time and historical financial data for equities — prices, news, financial statements, metrics, analyst estimates, insider/institutional activity, SEC filings, earnings press releases, segmented revenues, stock screening, and macro interest rates.
crypto-market-dataReal-time and historical cryptocurrency market data via CoinGecko: prices, market-cap rankings, charts, OHLC candles, token lookup, exchanges, categories, trending searches, and news.
prediction-market-dataPrediction markets data — Polymarket, Kalshi markets, prices, positions, and trades.
prediction-market-arbitrageFind and analyze arbitrage opportunities across prediction markets.
multi-source-searchIntelligent search for autonomous agents with structured retrieval plus Perplexity Sonar answer endpoints.
perplexity-searchPerplexity Sonar search and answer generation through AIsa.
youtube-serpYouTube SERP for autonomous agents. Search top-ranking videos, channels, and trends.
media-genGenerate images & videos with a single AIsa API key. 4 image models across 3 endpoints, 4 Wan video variants.
last30daysMulti-source research brief covering the last 30 days. Aggregates web, news, social, and financial signals into a single report.
aisa-providerConfigure OpenClaw to use AIsa as its model provider for Chinese and global LLM workflows.
aisa-tavilySearch the web and extract public page content for source discovery, current news lookup, and research workflows.
aisa-twitter-apiSearch, monitor, and post on X/Twitter through read, OAuth, and approved publishing workflows.
aisa-twitter-command-centerSearch X/Twitter profiles, timelines, tweets, trends, lists, communities, and Spaces for social research.
aisa-twitter-post-engageSearch X/Twitter and perform approved posting, liking, following, and unfollowing workflows.
aisa-youtube-searchSearch YouTube videos, channels, and playlists for video research and discovery.
cn-llmRoute Chinese-language prompts to Chinese LLMs such as Qwen, DeepSeek, GLM, and Baichuan.
last30days-zhResearch the last 30 days of signals across web, X, YouTube, Polymarket, and social sources in Chinese workflows.
llm-routerRoute prompts across GPT, Claude, Gemini, Qwen, DeepSeek, Grok, and other supported models.
marketRetrieve stock, crypto, financial statement, news, analyst, insider, and macro market data for analysis.
multi-searchSearch across web, academic, smart search, Tavily, and research-answer sources for stronger coverage.
perplexity-researchGenerate citation-backed research answers using Perplexity Sonar models.
prediction-market-arbitrage-zhFind cross-platform prediction market price differences and potential arbitrage opportunities.
prediction-market-data-zhQuery Polymarket and Kalshi prices, order books, positions, trades, and wallet metrics.
scholar-searchSearch academic papers and scholarly sources for research tasks.
smart-searchRun hybrid search that combines web and scholarly sources for broader research coverage.
stock-analysisAnalyze stocks and crypto assets with scoring, risk flags, entry levels, targets, and stop levels.
stock-dividendAnalyze dividend yield, payout safety, growth, income quality, and dividend track records.
stock-hotFind high-momentum stocks and crypto assets, market movers, catalysts, and watchlist candidates.
stock-portfolioTrack portfolio positions, allocation, and live profit and loss for stocks and crypto assets.
stock-rumorsScan for early market signals such as M&A rumors, insider activity, analyst changes, and regulatory news.
stock-watchlistManage a stock and crypto watchlist with price targets, stops, and alert checks.
tavily-extractExtract clean readable content from public URLs for article, document, and source analysis.
tavily-searchSearch the web with configurable depth, topic filters, time ranges, and domain filters.
twitter-command-center-search-postSearch X/Twitter and publish approved posts through OAuth-backed workflows.
us-stock-analystCreate US stock analysis reports using financial data, news, filings, sentiment, and AI synthesis.
web-searchSearch current web sources and return structured results with titles, URLs, and snippets.
x-intelligence-automationSearch X/Twitter and run approved engagement workflows for likes, follows, unfollows, and posts.
youtube-searchSearch YouTube videos, channels, and playlists for content research, trend discovery, competitor tracking, and topic exploration.
Showing 43 of 43 capabilities
How It Works
Your agent fetches the agent card from the well-known URL. It learns what AIsa can do, how to authenticate, and which skills are available.
The agent card declares Bearer token auth. Your agent includes the API key in the Authorization header — no OAuth flows needed.
Your agent selects the right skill based on tags and descriptions, constructs the API call using the OpenAPI spec, and executes autonomously.
Code Examples
# Fetch the agent card to discover AIsa's capabilities
curl -s https://aisa.one/.well-known/agent-card.json | jq .
# Response includes:
# - name, description, provider info
# - authentication requirements (Bearer token)
# - skills array with IDs, descriptions, tags, and examples
# - supported input/output modesimport httpx
# Step 1: Discover AIsa's capabilities via A2A agent card
agent_card = httpx.get("https://aisa.one/.well-known/agent-card.json").json()
print(f"Agent: {agent_card['name']}")
print(f"Auth: {agent_card['authentication']['schemes']}")
print(f"Skills: {len(agent_card['skills'])} available")
for skill in agent_card["skills"]:
print(f" - {skill['name']}: {skill['description'][:80]}...")
# Step 2: Use a discovered skill (e.g., chat completions)
response = httpx.post(
f"{agent_card['url']}/v1/chat/completions",
headers={"Authorization": "Bearer YOUR_AISA_API_KEY"},
json={
"model": "claude-sonnet-4-20250514",
"messages": [{"role": "user", "content": "Hello from an autonomous agent!"}]
}
)
print(response.json()["choices"][0]["message"]["content"])// Agent-to-Agent discovery and integration with AIsa
const AGENT_CARD_URL = "https://aisa.one/.well-known/agent-card.json";
interface AgentSkill {
id: string;
name: string;
description: string;
tags: string[];
examples: string[];
}
interface AgentCard {
name: string;
url: string;
authentication: { schemes: string[] };
skills: AgentSkill[];
}
// Step 1: Discover the agent's capabilities
const card: AgentCard = await fetch(AGENT_CARD_URL).then(r => r.json());
// Step 2: Find a skill by tag
const searchSkill = card.skills.find(s => s.tags.includes("web-search"));
if (searchSkill) {
console.log(`Found skill: ${searchSkill.name}`);
console.log(`Description: ${searchSkill.description}`);
}
// Step 3: Call the discovered API
const result = await fetch(`${card.url}/apis/v1/tavily/search`, {
method: "POST",
headers: {
"Authorization": "Bearer YOUR_AISA_API_KEY",
"Content-Type": "application/json"
},
body: JSON.stringify({
query: "latest AI agent frameworks 2026",
topic: "news",
max_results: 5
})
}).then(r => r.json());
console.log(result);Connect AIsa as an MCP server in Claude Desktop, Cursor, or any MCP-compatible client.
{
"mcpServers": {
"aisa": {
"command": "npx",
"args": ["-y", "@anthropic/mcp-remote", "https://aisa.one/docs/mcp"],
"env": {
"AISA_API_KEY": "your-api-key-here"
}
}
}
}Get your API key and start integrating AIsa into your agent workflows. Pay per call, no subscriptions.