api

All endpoints return JSON. No authentication required. Data is updated regularly and served via CDN.

Base URL: https://forgeuno.com

models

GET /api/v1/models.json

All LLM providers, their models, pricing, API configuration, and estimated monthly cost for agent workloads.

response structure

{
  "version": 6,
  "updated_at": "2026-03-20T00:00:00Z",
  "exchange_rates": { "USD_CNY": 7.2, "as_of": "2026-03-20" },
  "cost_tiers": {
    "agent": {
      "starter":  { "monthly_tasks": 40,  "llm_calls": 800,  ... },
      "regular":  { "monthly_tasks": 200, "llm_calls": 4000, ... },
      "power":    { "monthly_tasks": 600, "llm_calls": 12000, ... }
    }
  },
  "providers": [
    {
      "id": "anthropic",
      "display": { "en": { "name": "...", "description": "..." } },
      "website": "https://...",
      "docs_url": "https://...",
      "pricing_url": "https://...",
      "regions": ["us", "eu"],
      "api_type": "anthropic-messages",
      "base_url": "https://api.anthropic.com",
      "auth_mode": "api_key",
      "auth_keys": [{ "id": "ANTHROPIC_API_KEY", "description": "..." }],
      "pricing": "Pay-as-you-go per token. ...",
      "default_model": "claude-sonnet-4-6",
      "models": [
        {
          "id": "claude-sonnet-4-6",
          "display": { "en": { "name": "...", "description": "..." } },
          "context_window": 1000000,
          "max_output_tokens": 64000,
          "reasoning": true,
          "pricing": "Input $3/M, output $15/M. ...",
          "monthly_cost": {
            "agent": {
              "starter":  { "cost": 15.0, "currency": "USD", "cost_usd": 15.0, "cost_cny": 108.0, "calculation": "..." },
              "regular":  { "cost": 74.85, ... },
              "power":    { "cost": 224.55, ... }
            }
          },
          "compat": { "supports_streaming": true, "supports_tool_use": true, ... },
          "status": "ga"
        }
      ],
      "status": "active",
      "last_verified": "2026-03-19"
    }
  ]
}

key fields

field	description
`provider.api_type`	API protocol: `openai-completions`, `anthropic-messages`, `google-generative-ai`
`provider.auth_keys`	Required credentials. Each has an `id` (env var name) and `description`
`provider.regions`	Available regions: `us`, `eu`, `cn`
`model.monthly_cost`	Estimated monthly cost per usage tier, in original currency + USD/CNY conversions
`model.compat`	Feature support flags: streaming, tool use, vision, system prompt
`model.status`	`ga` (stable), `preview`, or `deprecated`
`cost_tiers`	Defines usage assumptions for monthly cost calculation (tasks, LLM calls, tokens per month)

tools

GET /api/v1/tools.json

MCP tools and external services with security assessments, integration guides, and capability descriptions.

response structure

{
  "version": 3,
  "updated_at": "2026-03-15T00:00:00Z",
  "tools": [
    {
      "id": "github-mcp",
      "name": "GitHub MCP Server",
      "category": "code",
      "description": "...",
      "integration": {
        "type": "mcp-server",
        "package": "@github/mcp-server",
        "install_command": "npx -y @github/mcp-server"
      },
      "security": {
        "level": "caution",
        "permissions": ["repo-read", "repo-write"],
        "data_exposure": "repository contents",
        "requires_supervisor": false
      },
      "capabilities": {
        "actions": ["read-files", "create-pr", ...],
        "when_to_use": "..."
      },
      "free_tier": true
    }
  ]
}

key fields

field	description
`security.level`	`safe` (autonomous use), `caution` (log usage), `restricted` (requires approval)
`integration.type`	`mcp-server`, `api`, or `sdk`
`capabilities.actions`	What the tool can do
`capabilities.when_to_use`	Natural language guidance for agents

skills

GET /api/v1/skills.json

Installable agent skills — workflow templates that give agents domain-specific capabilities.

response structure

{
  "version": 2,
  "updated_at": "2026-03-17T00:00:00Z",
  "skills": [
    {
      "id": "industry-briefing",
      "name": "Industry Briefing",
      "category": "writing",
      "description": "...",
      "install": {
        "type": "download",
        "source": "official",
        "skill_md_url": "https://forgeuno.com/skills/industry-briefing/SKILL.md"
      },
      "security": {
        "level": "safe",
        "permissions": [],
        "data_exposure": "none",
        "requires_supervisor": false
      },
      "capabilities": {
        "triggers": ["scheduled", "on-demand"],
        "workflows": ["research", "write", "deliver"],
        "when_to_use": "..."
      },
      "listing_status": "active"
    }
  ]
}

key fields

field	description
`install.type`	`download` — download skill files (SKILL.md + optional references/) to agent's skill directory
`install.skill_md_url`	Direct URL to the skill definition file
`listing_status`	`active` skills are ready to install; others are drafts

All responses are static JSON files served via CDN. Cache headers allow aggressive caching. Rate limiting is not applied.