{"name":"Strata MCP","description":"AI ecosystem intelligence for agents. Best practices, news, integrations, and search across 22+ AI ecosystems. Call list_ecosystems first to discover available slugs for your tier.","version":"1.0.0","tools":[{"name":"get_best_practices","description":"Get AI-verified best practices for an AI ecosystem. Each result includes: title, body, source_urls[], confidence (high/medium/low), content_age_hours, data_freshness. Results ordered by recency. Check data_freshness before acting on time-sensitive guidance. Strata provides intelligence, not ground truth. Always verify critical decisions against the source_urls returned with each item."},{"name":"get_latest_news","description":"Get the latest news for an AI ecosystem. Pro tier: real-time results. Free tier: items older than 24h. Each result includes: title, body, published_at, source_urls[], content_age_hours, data_freshness. Strata provides intelligence, not ground truth. Always verify critical decisions against the source_urls returned with each item."},{"name":"get_top_integrations","description":"Get integrations and MCP servers for an AI ecosystem. Without use_case: returns all integrations ordered by recency. With use_case: returns results ranked by full-text relevance. Each result includes: title, body, source_urls[]; when use_case is provided, also includes rank (relevance position). Quarantined items are never returned. Example: get_top_integrations(ecosystem=\"claude\", use_case=\"code review\") finds tools Claude users use for code review. Returns curated documentation and integration guides — not MCP servers. Use find_mcp_servers to discover connectable MCP protocol servers. Strata provides intelligence, not ground truth. Always verify critical decisions against the source_urls returned with each item."},{"name":"search_ecosystem","description":"Full-text search across verified AI ecosystem content — best practices, news, integrations, and security advisories. Includes real-time web results when the corpus confidence is low, making this the right tool for open-ended security research, vulnerability lookups, and queries about recent events. Each result includes a security envelope (risk_level, trust_score, verification_status). Use this when get_threat_feed returns empty or when you need broader context beyond known risk-signal changes. Use find_mcp_servers to search the MCP server directory by capability. Strata provides intelligence, not ground truth. Always verify critical decisions against the source_urls returned with each item."},{"name":"list_ecosystems","description":"List all AI ecosystems available on your current tier. Returns your_tier and an ecosystems array with slug, name, and vendor per entry. Call this first to get valid slugs before using get_best_practices, get_latest_news, get_top_integrations, or search_ecosystem."},{"name":"find_mcp_servers","description":"Search 2,100+ MCP servers by use case or keyword using semantic similarity weighted by trust scores. Each result includes: security_score (0–100, repo health), runtime_score (0–100, tool behavior analysis), capability_flags (e.g. \"shell_exec\", \"fs_write\", \"dynamic_eval\"), hosted_endpoint, tool_count, runtime_freshness (fresh/aging/stale/unknown), circuit_broken (true if Strata has automatically tripped a circuit breaker due to a critical risk event — skip these servers or pass exclude_circuit_broken=true). Quarantined and archived servers are always excluded. Strata provides intelligence, not ground truth. Always verify critical decisions against the source_urls returned with each item."},{"name":"verify_payment_endpoint","description":"Verifies an x402 payment endpoint before an agent pays it. Checks SSL validity, domain age, payment amount reasonableness, and whether the 402 response is well-formed. Returns a trust score (0-100), risk level (critical/high/medium/low), and capability flags. Use this before any autonomous payment to confirm the endpoint is legitimate. Does not make any actual payment. Returns cached results if checked within 24 hours."},{"name":"track_data_flow","description":"Records a data flow between two MCP servers for lineage tracking. Call this after your agent reads data from one server and sends it to another — it creates an auditable record of where your data traveled and flags if the destination has net_egress or other risky capabilities. The flow is immediately visible in the Strata dashboard under Data Lineage."},{"name":"verify_agent_credential","description":"Verifies a Strata-issued agent credential (JWT). MCP servers and x402 endpoints call this to confirm an agent is who it claims to be before honouring its tool calls. Returns the agent ID, owning profile, declared capabilities, expiration, and live revocation status. Use this for high-trust calls (writes, payments) where instant revocation matters; for low-stakes reads, prefer offline JWKS verification at /.well-known/jwks.json."},{"name":"get_threat_feed","description":"Returns recent risk-signal changes for MCP servers — quarantines, new dangerous capability flags, security score drops, and injection detections. Use this to check whether any server you rely on has changed risk profile recently. Set affected_only=true to filter to servers you have actually connected to. When the feed returns empty, follow the next_step field in the response."}],"prompts":[{"name":"ecosystem_briefing","description":"Generate a structured intelligence briefing for any AI ecosystem using Strata tools","arguments":[{"name":"ecosystem","description":"The ecosystem to brief on: claude, openai, gemini, langchain, ollama, cursor, groq, etc.","required":true}]},{"name":"cross_ecosystem_compare","description":"Compare two AI ecosystems side by side using Strata intelligence","arguments":[{"name":"ecosystem_a","description":"First ecosystem to compare","required":true},{"name":"ecosystem_b","description":"Second ecosystem to compare","required":true}]},{"name":"agent_stack_review","description":"Stack recommendation for a specific use case based on current Strata intelligence","arguments":[{"name":"use_case","description":"What you are building e.g. \"RAG pipeline\", \"coding assistant\", \"data analysis agent\"","required":true}]}],"resources":[{"uri":"strata://formatting-guide","name":"Strata Formatting Guide","description":"Instructions for presenting Strata intelligence results"}],"auth":"Pass your Strata API key as: Authorization: Bearer sk_your_key or X-API-Key: sk_your_key","endpoint":"https://usestrata.dev/mcp"}