AI agent intelligence

Give your AI agents live market & security intelligence

A 14-tool MCP server, a commercial REST API, SSE alert streams, and a machine-readable dossier on every trend — so the agents working for you stay current without scraping.

Mint an API keyRead the docs
MCP tools
14
MCP tools
Streamable HTTP, JSON-RPC
Auth headers
2
Auth headers
X-MCP-Key or Bearer vel_*
Commercial API
REST
Commercial API
/api/v1/trends + /pulse
Categories
16
Categories
Trends to security to ventures
Agent-native by design

Intelligence your agents can query, not scrape

Velocity assumes the consumer might be a model — every surface is structured.

Pointing an agent at a news site means brittle scraping, parsing HTML, and re-deriving what matters. Velocity removes that work: the same intelligence a human reads on the dashboard is available to agents through a Model Context Protocol (MCP) server, a commercial REST API, and a stable, versioned AgentDossierSummary on every active trend page.

Connect from Claude Code, ChatGPT Developer Mode, Codex, OpenClaw, or the OpenAI Responses API. Authenticate with a vel_* key — as X-MCP-Key or Authorization: Bearer — and your agent can search trends, brief itself, analyze momentum, and watch your stack for new CVEs.

The toolbelt

14 read-only MCP tools

Grouped by what an agent does. PRO+ tools are marked; the rest are available on every tier.

Trend search & retrieval

  • search

    Canonical retrieval — search active trends, return id/title/url/text/metadata (works in ChatGPT & Claude remote-MCP).

  • fetch

    Canonical retrieval — fetch one trend by id with full text and metadata.

  • velocity_search_trends

    Search trends by keyword, category, or velocity score.

  • velocity_trend_details

    Full detail plus recent velocity history for a single trend.

  • velocity_intel_brief

    Top and emerging trends plus a category breakdown for a time window.

Signal analysis

  • velocity_momentum_checkPRO+

    Whether a trend is accelerating, plateauing, or fading.

  • velocity_correlation_lookupPRO+

    Pearson correlation of a trend’s velocity history against others.

  • velocity_watchlist_alerts

    List threshold breaches on the caller’s own watchlists.

Security & threat intelligence

  • security_vuln_lookup

    Look up a CVE or GHSA advisory and its details.

  • security_build_stack_from_file

    Parse a dependency / lockfile into a watched tech stack.

  • security_list_stack_watch

    List the tech stack the caller is watching.

  • security_subscribe_stack_hitsPRO+

    Subscribe so the agent is alerted when a new CVE matches the watched stack.

  • threat_actor_lookup

    Look up an APT / threat-actor group, aliases, origin, motivation, and TTPs.

  • threat_actor_search

    Search 200+ threat actors from MITRE ATT&CK, MISP, and CISA.

Want the full schema, response shapes, and setup snippets? See the MCP server reference →

Use it for

What an agent can do with Velocity

Six patterns teams wire in on day one.

Quickstart

Two ways to connect

Add the MCP server to your agent, or poll the REST API from any pipeline.

MCP server (Claude Code)

claude mcp addbash
# Connect Claude Code to the Velocity MCP server
claude mcp add --transport http \
  --header "X-MCP-Key: vel_your_api_key" \
  velocity https://velocity.andiamo.tech/api/agents/mcp

Commercial REST API

GET /api/v1/trendsbash
# Poll the commercial REST API (Bearer vel_* key, per-tier daily quota)
curl -H "Authorization: Bearer vel_your_api_key" \
  "https://velocity.andiamo.tech/api/v1/trends?category=security&minScore=50&limit=10"
Compatible with

Connect from your agent runtime

Streamable HTTP MCP plus canonical search/fetch for the OpenAI ecosystem.

FAQ

Agent integration questions

  • How do AI agents connect to Velocity?

    Two ways. Most agents use the Model Context Protocol (MCP) server at /api/agents/mcp over Streamable HTTP — Claude Code, ChatGPT Developer Mode, Codex, OpenClaw, and the OpenAI Responses API all connect to it. For non-MCP pipelines there is a commercial REST API: GET /api/v1/trends and /api/v1/pulse with a Bearer vel_* key.

  • How is the agent authenticated?

    With a vel_* API key you mint in the Agent Hub. The MCP server accepts it as X-MCP-Key (Claude Code / OpenClaw native) or as Authorization: Bearer (Codex, OpenAI Responses, ChatGPT). Velocity stores only a SHA-256 hash of the key, never the plaintext, and every tool call is authorized as the issuing user. The REST API uses the same Bearer vel_* scheme.

  • What can an agent actually do with the 14 tools?

    Search and fetch trends, pull an intel brief for a window, run momentum and correlation checks (PRO+), list watchlist alerts, look up CVEs and parse a lockfile into a watched stack, subscribe to alerts when a new CVE matches that stack (PRO+), and look up or search 200+ threat actors. All tools are read-only.

  • Can my agent get alerted when a new CVE drops for my stack?

    Yes. Use security_build_stack_from_file to register the software and versions you run, then security_subscribe_stack_hits (PRO+) so the agent is notified the moment a new CVE or GitHub Security Advisory matches the watched stack. Alerts can also flow via email, push, webhook, and a Server-Sent-Events stream.

  • Is there a usage limit?

    Tool calls and REST requests are rate-limited per user by subscription tier, with daily quotas. Exceeding the limit returns a 429 with a Retry-After header (or a JSON-RPC rate-limit error over MCP). Higher tiers raise both the per-minute and per-day ceilings.

  • Can I embed Velocity data in my own product?

    Commercial use — embedding Velocity data in a product you ship — is permitted on PRO and above. The FREE tier can view all trends and use the tools for personal and internal work.

Wire Velocity into your agents

Mint a vel_* key in the Agent Hub, connect the MCP server or the REST API, and your agents are reading scored intelligence in minutes.

Open the Agent HubSee the MCP server