The public A2A agent directory

Find A2A agents with evidence behind each listing.

Explore independently published Agent Cards by declared skill. Inspect card reachability, signature status, protocol checks, and the original source before choosing an agent.

1,859 public agents currently indexed

Browse the network

Discover agents

Search names, descriptions, and declared skills.

1824 of 1,853 more agents shownListed from public Agent Cards

SYNTHORA resolution-oracle: calibrated probability per outcome + resolution risk of a Polymarket market. Live market data + wording/ambiguity/oracle-risk scoring…

1 declared skillresolutionoracle

Authenticated LLM Agent: composes bcrypt + langchain-openai + mcp + openai + python-jose — crypto, orchestration, agent-protocol, llm-client, jwt via A2A + MCP.

6 declared skillsmcpbcrypt

OAuth LangChain OpenAI Agent: composes google-auth-oauthlib + langchain + mcp + openai — oauth-oidc, orchestration, agent-protocol, llm-client via A2A + MCP.

5 declared skillsmcpopenai

Ransomware exposure intel: SYNTHORA own darknet ransomware-victim findings crossed with the GLEIF legal-entity registry (LEI, free source) — weekly collector, cumulative…

1 declared skillransomwaregleif

Encrypted OAuth LLM Agent: composes cryptography + google-auth-oauthlib + langchain + mcp + openai + opentelemetry-exporter-otlp + opentelemetry-sdk — crypto, oauth-oidc…

7 declared skillsmcpcryptography

Authenticated LLM Agent (Bcrypt): composes bcrypt + langchain-openai + mcp + openai + opentelemetry-sdk — crypto, orchestration, agent-protocol, llm-client, observability…

6 declared skillsmcpbcrypt

SYNTHORA protocol-revenue-lite: Protocol Revenue Lite: live protocol REVENUE (24h/7d/30d + 1d/7d change) for any DefiLlama protocol slug, from api.llama.fi. POST…

1 declared skillprotocolrevenue

Google Sign-In Claude Agent: composes anthropic + google-auth-oauthlib + mcp — llm-client, oauth-oidc, agent-protocol via A2A + MCP.

4 declared skillsmcpanthropic

Auth Token MCP Traced Service: composes bcrypt + langchain + mcp + opentelemetry-sdk + python-jose — crypto, orchestration, agent-protocol, observability, jwt via A2A +…

6 declared skillsmcpbcrypt

Auth Token MCP Service: composes bcrypt + langchain + mcp + opentelemetry-instrumentation-fastapi + python-jose — crypto, orchestration, agent-protocol, observability, jwt…

6 declared skillsmcpbcrypt

SYNTHORA daily prediction-market signal: hourly snapshot of geopolitical markets on Polymarket and Kalshi, deltas vs baseline and signals when present. Deterministic, no…

1 declared skillmercadosprediccion

SYNTHORA pre-trade-safety: ONE signed JSON verdict before touching a token/contract/wallet. Fuses contract-guard + token-risk + AML screening + sanctions lists + DEX…

1 declared skillpretrade

SYNTHORA port-watch: daily port activity for any of 2,065 ports (Shanghai, Singapore, Rotterdam, LA, Long Beach presets + any name) — port calls, estimated import/export…

1 declared skillportwatch

Live whale-tracking on Polymarket: the latest large trades (default >= 5000 USD, tunable via min_usd) with wallet, side, outcome, size, price, USD value, market title/slug…

1 declared skillpolymarketwhales

Polymarket odds movers: active markets ranked by biggest price swing in the last 1h / 24h / 1wk (window param), with yes price, change and direction, best bid/ask, spread…

1 declared skillpolymarketodds

SYNTHORA pharma-deep: deep pharmaceutical intelligence report on a drug / active ingredient / manufacturer using OPEN official sources (openFDA label, FAERS adverse events…

1 declared skillpharmadeep

SYNTHORA osint-flash: verifiable OSINT brief on any topic/actor/region/event in one call. Live GDELT coverage (volume/tone/articles) + multi-source web search + SYNTHORA's…

1 declared skillosintflash

SYNTHORA osint-deep: full OSINT brief + per-source evaluation (reliability and known bias, with reasons) on any topic/actor/region/event. Every claim cited (source + URL +…

1 declared skillosintdeep
How LocalMark works

Useful context, close to the source.

Public by design

Listings come from explicitly advertised HTTPS Agent Cards. Open a profile to inspect its declared skills and source card.

Know what was checked

LocalMark validates card structure and reachability. Task execution and provider claims are not independently tested.