Documentation

Build with Argosor

Everything in the console is available to your code and your agents: 15 MCP tools over Streamable HTTP, and a REST API under /v1.

Run an AI visibility audit

curl https://argosor.com/v1/geo/audit \
  -H "Authorization: Bearer arg_live_your_key" \
  -H "Content-Type: application/json" \
  -d '{"brandName":"Acme","domain":"acme.io","promptTopic":"best roadmap tool for startups"}'

Connect an MCP client

The endpoint is https://argosor.com/mcp (protocol 2025-03-26). initialize and tools/list are public, and a tool call without a key provisions a sandbox workspace automatically.

CursorAdd to Cursor →

~/.cursor/mcp.json

{
  "mcpServers": {
    "argosor": {
      "url": "https://argosor.com/mcp",
      "headers": {
        "Authorization": "Bearer arg_live_your_key"
      }
    }
  }
}

Claude Code

Terminal

claude mcp add --transport http argosor https://argosor.com/mcp \
  --header "Authorization: Bearer arg_live_your_key"

Claude Desktop

claude_desktop_config.json

{
  "mcpServers": {
    "argosor": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://argosor.com/mcp"
      ]
    }
  }
}

Windsurf

mcp_config.json

{
  "mcpServers": {
    "argosor": {
      "serverUrl": "https://argosor.com/mcp",
      "headers": {
        "Authorization": "Bearer arg_live_your_key"
      }
    }
  }
}

ChatGPT

MCP server URL

https://argosor.com/mcp

VS CodeInstall in VS Code →

Terminal

code --add-mcp '{"name":"argosor","type":"http","url":"https://argosor.com/mcp"}'

OpenClaw

Message

Install the Argosor skill from https://argosor.com/.well-known/agent-skills/argosor-mcp/SKILL.md
and connect to the MCP server at https://argosor.com/mcp.

Authentication

Send a workspace key as a bearer token: Authorization: Bearer arg_live_…. Create keys in the console under Integrations, or with POST /v1/keys. Session tokens (arg_sess_…) issued at sign-in work the same way.

Create a workspace from the command line

curl https://argosor.com/v1/workspaces \
  -H "Content-Type: application/json" \
  -d '{"workspaceName":"Acme Growth","email":"you@acme.io","password":"a-strong-password"}'

REST API

JSON in, JSON out. The full schema lives in /openapi.json.

Method	Path	What it does
POST	/v1/workspaces	Create a workspace and its first API key
GET	/v1/workspaces	Return the workspace that owns the token
POST	/v1/auth/login	Sign in with email + password or an API key
GET	/v1/account	Plan, usage, API keys (masked) and BYOK vault
POST	/v1/keys	Create a scoped API key
POST	/v1/byok	Store an encrypted provider key
POST	/v1/geo/audit	AI visibility audit across nine models
POST	/v1/geo/reputation-writer	7-dimension reputation score + answer-first writer pack
POST	/v1/geo/agent-attribution	GA4, Plausible and PostHog AI traffic attribution
POST	/v1/serp/search	Live SERP with answer summary and People Also Ask
POST	/v1/markdown/extract	Clean Markdown, token count and SHA-256 for any URL
POST	/v1/keywords/cluster	Keyword ideas grouped by intent
POST	/v1/backlinks/authority	Authority index, referring domains and anchors
POST	/v1/competitors/gap	Keyword and citation gaps against competitors
POST	/v1/schema/generate	Schema.org JSON-LD for any URL
POST	/v1/seo/crawl	Technical SEO, crawler and firewall probe
POST	/v1/seo/cms-autopilot	Action Board, staging, approve & publish, rollback
POST	/v1/billing/checkout	Start a plan upgrade
DELETE	/v1/account	Delete the workspace and its vault

MCP tools

15 tools, each with a strict JSON Schema and behavior annotations. Expand a tool to see its input.

serp_live_search

Live Multi-Engine SERP & AI Overview Search

Read-only

Execute a real-time Search Engine Results Page (SERP) query across Google, Bing, DuckDuckGo, Google News, or Google AI Overviews. Returns ranked organic listings, AI Overview synthesis and cited URLs, People Also Ask (PAA) questions, and traffic estimates. Supports Managed Edge execution ($0.0005/query, 70% cheaper than legacy SERP APIs) or BYOK ($0 markup).

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "query"
  ],
  "properties": {
    "query": {
      "type": "string",
      "description": "Search query or keyword phrase to inspect in real time (e.g., \"best seo api for ai agents\")."
    },
    "engine": {
      "type": "string",
      "enum": [
        "google",
        "bing",
        "duckduckgo",
        "google_news",
        "google_ai_overview"
      ],
      "description": "Target search engine vertical to query. Defaults to \"google\"."
    },
    "country": {
      "type": "string",
      "description": "Two-letter ISO 3166-1 alpha-2 country code for localized geo-targeting (e.g., \"us\", \"gb\", \"de\", \"es\"). Defaults to \"us\"."
    },
    "language": {
      "type": "string",
      "description": "Two-letter ISO 639-1 language code for SERP localization (e.g., \"en\", \"es\", \"fr\"). Defaults to \"en\"."
    },
    "limit": {
      "type": "integer",
      "minimum": 1,
      "maximum": 30,
      "description": "Maximum number of organic SERP results to return (1 to 30). Defaults to 10."
    },
    "includeAiOverview": {
      "type": "boolean",
      "description": "When true, extracts Google AI Overview / generative summary and cited source URLs. Defaults to true."
    },
    "includePeopleAlsoAsk": {
      "type": "boolean",
      "description": "When true, extracts People Also Ask (PAA) questions and answer snippets. Defaults to true."
    },
    "executionMode": {
      "type": "string",
      "enum": [
        "managed",
        "byok"
      ],
      "description": "Execution mode: \"managed\" ($0.0005/query via Argosor Global Edge) or \"byok\" ($0 markup using encrypted vault credentials). Defaults to \"managed\"."
    }
  }
}

geo_llm_visibility_audit

GEO & Multi-LLM Citation Visibility Audit

Read-only

Run a Generative Engine Optimization (GEO) and LLM Citation Share-of-Voice (SoV) audit across ChatGPT Search, Perplexity, Claude, Google AI Overviews, Gemini, and Grok. Probes live domain signals (/llms.txt, Schema.org JSON-LD) and returns brand mention rate, citation rank, competitor displacement, and concrete recommendations to capture more AI citations.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "brandName",
    "domain"
  ],
  "properties": {
    "brandName": {
      "type": "string",
      "description": "Target brand or product name to audit across generative engines (e.g., \"Argosor\")."
    },
    "domain": {
      "type": "string",
      "description": "Primary canonical domain of the brand being audited (e.g., \"argosor.com\")."
    },
    "promptTopic": {
      "type": "string",
      "description": "High-intent buyer prompt or category topic to evaluate Share-of-Voice for (e.g., \"best seo and geo api for ai agents\")."
    },
    "competitors": {
      "type": "array",
      "items": {
        "type": "string",
        "description": "Competitor domain name (e.g., \"dataforseo.com\", \"serpapi.com\")."
      },
      "description": "Optional list of competitor domains to benchmark citation displacement against."
    },
    "llmEngines": {
      "type": "array",
      "items": {
        "type": "string",
        "enum": [
          "chatgpt_search",
          "perplexity",
          "claude",
          "google_ai_overviews",
          "google_ai_mode",
          "gemini",
          "grok",
          "deepseek",
          "copilot",
          "meta_ai"
        ],
        "description": "Identifier of the generative AI search engine to audit."
      },
      "description": "Optional subset of the 9 generative AI engines to include in the Share-of-Voice audit."
    },
    "forceRefresh": {
      "type": "boolean",
      "description": "When true, bypasses the 6-hour response deduplication cache and forces a fresh multi-engine audit. Defaults to false."
    }
  }
}

technical_seo_crawl

Live Technical SEO & AI Crawler Readiness Audit

Read-only

Perform a live HTTP + HTML technical SEO and AI-crawler readiness audit on any URL. Inspects HTTP status, response latency, title, meta description, canonical tag, H1-H3 heading hierarchy, Schema.org JSON-LD blocks, robots.txt permissions for AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended), and /llms.txt availability.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "url"
  ],
  "properties": {
    "url": {
      "type": "string",
      "description": "Full URL or domain to crawl and audit (e.g., \"https://argosor.com\")."
    },
    "checkAiCrawlers": {
      "type": "boolean",
      "description": "When true, fetches /robots.txt on the target origin to verify whether GPTBot, ClaudeBot, and PerplexityBot are permitted. Defaults to true."
    },
    "checkLlmsTxt": {
      "type": "boolean",
      "description": "When true, probes /llms.txt on the target origin for machine-readable agent documentation. Defaults to true."
    },
    "checkSchemaOrg": {
      "type": "boolean",
      "description": "When true, parses and counts embedded application/ld+json Schema.org blocks. Defaults to true."
    }
  }
}

keyword_intent_cluster

Semantic Keyword Intent & Pillar Clustering

Read-only

Generate semantic keyword clusters with monthly search volume, CPC (USD), Keyword Difficulty (KD 0-100), buyer intent classification (commercial, transactional, informational, navigational), and recommended pillar-and-spoke URL slugs for both organic search and LLM citation capture.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "seedTopic"
  ],
  "properties": {
    "seedTopic": {
      "type": "string",
      "description": "Core seed keyword or product category to expand into semantic intent clusters (e.g., \"generative engine optimization api\")."
    },
    "country": {
      "type": "string",
      "description": "Two-letter ISO country code for search volume and CPC calibration (e.g., \"us\", \"gb\"). Defaults to \"us\"."
    },
    "language": {
      "type": "string",
      "description": "Two-letter ISO language code for keyword generation (e.g., \"en\"). Defaults to \"en\"."
    },
    "maxClusters": {
      "type": "integer",
      "minimum": 1,
      "maximum": 10,
      "description": "Maximum number of semantic topic clusters to generate (1 to 10). Defaults to 4."
    }
  }
}

backlink_authority_graph

Backlink Authority & LLM Citation Graph

Read-only

Inspect a domain’s backlink authority profile and LLM citation graph. Returns Domain Rating (DR 0-100), referring domains count, total backlinks, dofollow/nofollow ratio, spam risk score, anchor text distribution, and top pages most frequently cited by AI search engines.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "domain"
  ],
  "properties": {
    "domain": {
      "type": "string",
      "description": "Target root domain or hostname to analyze (e.g., \"argosor.com\")."
    },
    "includeAnchorDistribution": {
      "type": "boolean",
      "description": "When true, includes categorized anchor text percentage breakdown. Defaults to true."
    },
    "includeLlmCitedPages": {
      "type": "boolean",
      "description": "When true, includes the top URLs on the domain most frequently cited by Perplexity, ChatGPT Search, and Claude. Defaults to true."
    }
  }
}

markdown_clean_extract

Live Web URL to Clean LLM Markdown Extractor

Read-only

Fetch any live web URL and convert it into clean, boilerplate-free, LLM-ready Markdown for RAG pipelines and autonomous research agents. Strips scripts, styles, ads, and navigation chrome while preserving heading hierarchy, lists, canonical links, token count, and SHA-256 content hash at 90% lower cost than legacy scrapers.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "url"
  ],
  "properties": {
    "url": {
      "type": "string",
      "description": "Full HTTP/HTTPS web page URL to fetch and convert into clean Markdown."
    },
    "includeLinks": {
      "type": "boolean",
      "description": "When true, extracts a structured list of outbound and internal anchor links from the page. Defaults to true."
    },
    "includeMetadata": {
      "type": "boolean",
      "description": "When true, includes title, word count, estimated LLM token count, and SHA-256 content digest. Defaults to true."
    },
    "maxCharacters": {
      "type": "integer",
      "minimum": 500,
      "maximum": 100000,
      "description": "Maximum character length of the returned Markdown document (500 to 100,000). Defaults to 16000."
    }
  }
}

competitor_gap_matrix

Multi-Domain SEO & GEO Competitor Gap Matrix

Read-only

Build a multi-domain SEO and GEO competitive gap matrix comparing a primary domain against up to 5 competitors. Identifies high-volume, low-difficulty keyword gaps, AI Share-of-Voice disparities, unit cost benchmarks, and the highest-ROI comparison pages (/vs/*) to publish next.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "primaryDomain",
    "competitorDomains"
  ],
  "properties": {
    "primaryDomain": {
      "type": "string",
      "description": "Your primary domain to benchmark (e.g., \"argosor.com\")."
    },
    "competitorDomains": {
      "type": "array",
      "items": {
        "type": "string",
        "description": "Competitor root domain (e.g., \"dataforseo.com\", \"serpapi.com\", \"ahrefs.com\", \"firecrawl.dev\")."
      },
      "description": "Array of 1 to 5 competitor domains to compare against the primary domain."
    },
    "focusVertical": {
      "type": "string",
      "description": "Optional market vertical or product category description for tailored gap prioritization."
    }
  }
}

schema_jsonld_generator

Deterministic Schema.org JSON-LD Graph Generator

Read-only

Generate a validated, deterministic Schema.org JSON-LD @graph payload (SoftwareApplication, Organization, FAQPage, Article, Product, WebSite) ready to embed in HTML <head>. Maximizes eligibility for Google Rich Results and improves entity disambiguation inside LLM knowledge graphs.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "entityType",
    "name",
    "url",
    "description"
  ],
  "properties": {
    "entityType": {
      "type": "string",
      "enum": [
        "SoftwareApplication",
        "Organization",
        "FAQPage",
        "Article",
        "Product",
        "WebSite"
      ],
      "description": "Primary Schema.org entity type to generate."
    },
    "name": {
      "type": "string",
      "description": "Canonical brand, product, or article title (e.g., \"Argosor\")."
    },
    "url": {
      "type": "string",
      "description": "Canonical URL for the entity (e.g., \"https://argosor.com\")."
    },
    "description": {
      "type": "string",
      "description": "Concise factual summary of the entity optimized for search snippets and LLM citations."
    },
    "pricingOffers": {
      "type": "array",
      "items": {
        "type": "object",
        "additionalProperties": false,
        "required": [
          "name",
          "priceUsd"
        ],
        "properties": {
          "name": {
            "type": "string",
            "description": "Plan or offer tier name (e.g., \"Starter Plan\")."
          },
          "priceUsd": {
            "type": "number",
            "description": "Monthly or unit price in USD (e.g., 19)."
          }
        }
      },
      "description": "Optional array of pricing tiers to embed as Schema.org Offer nodes."
    },
    "faqItems": {
      "type": "array",
      "items": {
        "type": "object",
        "additionalProperties": false,
        "required": [
          "question",
          "answer"
        ],
        "properties": {
          "question": {
            "type": "string",
            "description": "Frequently asked question text."
          },
          "answer": {
            "type": "string",
            "description": "Authoritative answer text."
          }
        }
      },
      "description": "Optional FAQ items to embed as a linked Schema.org FAQPage node."
    }
  }
}

account_register

Zero-Browser Agent Workspace Registration

Action

Zero-browser autonomous workspace registration. Creates a new Argosor account, relational workspace, and scoped API key (arg_live_...) in a single call, returning a magic console claim URL and $2.50 in free sandbox credits.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "email": {
      "type": "string",
      "description": "Optional developer or agent owner email address. If omitted, provisions an instant claimable sandbox identity."
    },
    "name": {
      "type": "string",
      "description": "Optional display name for the developer or autonomous agent."
    },
    "workspaceName": {
      "type": "string",
      "description": "Optional workspace name (e.g., \"Acme Growth & GEO Agent\")."
    }
  }
}

account_status

Workspace Quota, BYOK Vault & Billing Status

Read-only

Inspect the active Argosor workspace, plan tier (free, starter, builder, scale), remaining monthly query quota, USD credit ledger balance, minted API keys, and connected AES-256-GCM BYOK credentials.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "includeApiKeys": {
      "type": "boolean",
      "description": "When true, lists active API keys belonging to the workspace. Defaults to true."
    },
    "includeByokVault": {
      "type": "boolean",
      "description": "When true, lists encrypted BYOK provider credentials (DataForSEO, SerpApi, Ahrefs, Firecrawl, OpenAI, Perplexity) configured in the workspace vault. Defaults to true."
    }
  }
}

billing_upgrade

Stripe Live Plan Upgrade & Checkout Session

Action

Generate a Stripe Live Checkout session to upgrade the workspace plan (Starter $19/mo, Builder $39/mo, Scale $79/mo) or retrieve the Stripe Live catalog and Customer Billing Portal URL.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "plan"
  ],
  "properties": {
    "plan": {
      "type": "string",
      "enum": [
        "starter",
        "builder",
        "scale"
      ],
      "description": "Target Argosor subscription plan tier: \"starter\" ($19/mo), \"builder\" ($39/mo), or \"scale\" ($79/mo)."
    },
    "successUrl": {
      "type": "string",
      "description": "Optional redirect URL after completed Stripe Checkout payment."
    },
    "cancelUrl": {
      "type": "string",
      "description": "Optional redirect URL if checkout is canceled."
    }
  }
}

traffic_analytics

Admin Traffic & MCP Attribution Telemetry

Read-only

Query real-time web, REST API, and MCP JSON-RPC traffic attribution telemetry across acquisition channels, AI agent harnesses (Cursor, Claude, Grok, OpenClaw), MCP tools, and countries. Restricted to Allometrik master admin credentials.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "windowHours": {
      "type": "integer",
      "minimum": 1,
      "maximum": 720,
      "description": "Lookback time window in hours (1 to 720). Defaults to 24."
    },
    "limit": {
      "type": "integer",
      "minimum": 5,
      "maximum": 100,
      "description": "Maximum number of recent live events to include (5 to 100). Defaults to 25."
    }
  }
}

geo_reputation_and_writer

7D AI Brand Reputation Audit & Answer-First GEO Writer

Read-only

Run a 7-Dimension AI Brand Reputation & Narrative Audit (authority, trust & compliance, pricing value, technical depth, citation freshness, sentiment polarity, entity clarity) across live web, news, Hacker News, Wikipedia, and connected LLMs, and generate an Answer-First GEO Content & Markdown Twin Writer pack (Answer-First Markdown section, HTML comparison <table>, FAQPage JSON-LD, and Content-Signal robots.txt rule). Includes 6-hour response caching.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "brandName",
    "domain",
    "targetPrompt"
  ],
  "properties": {
    "brandName": {
      "type": "string",
      "description": "Brand or product name to evaluate and write Answer-First GEO content for (e.g., \"Argosor\")."
    },
    "domain": {
      "type": "string",
      "description": "Primary canonical domain of the brand (e.g., \"argosor.com\")."
    },
    "targetPrompt": {
      "type": "string",
      "description": "High-intent buyer prompt or topic to optimize citation capture for (e.g., \"best enterprise seo and geo mcp server\")."
    },
    "targetUrl": {
      "type": "string",
      "description": "Optional specific page URL being optimized (defaults to https://<domain>)."
    },
    "competitors": {
      "type": "array",
      "items": {
        "type": "string",
        "description": "Competitor root domain (e.g., \"llmpulse.ai\", \"dataforseo.com\")."
      },
      "description": "Optional list of up to 4 competitor domains for the generated Answer-First comparison table."
    },
    "forceRefresh": {
      "type": "boolean",
      "description": "When true, bypasses the 6-hour response deduplication cache. Defaults to false."
    }
  }
}

agent_crawler_traffic_attribution

3-Tier AI Crawler Readiness & GA4/Plausible/PostHog Attribution

Read-only

Audit a domain’s 3-tier AI crawler readiness—separating Training Crawlers (GPTBot, CCBot, Google-Extended), AI Search Indexers (OAI-SearchBot, PerplexityBot, Applebot-Extended), and Live User-Triggered Agent Browsers (ChatGPT-User, Claude-Web, Perplexity-User)—verify edge WAF bot-spoofing behavior, and generate ready-to-paste GA4, Plausible, PostHog, and Nginx AI traffic attribution regex filters covering 15+ AI answer engines.

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "domain"
  ],
  "properties": {
    "domain": {
      "type": "string",
      "description": "Target root domain to audit AI crawler taxonomy and generate analytics attribution filters for (e.g., \"argosor.com\")."
    },
    "analyticsPlatform": {
      "type": "string",
      "enum": [
        "all",
        "ga4",
        "plausible",
        "posthog"
      ],
      "description": "Target analytics platform for attribution snippets. Defaults to \"all\"."
    }
  }
}

cms_autopilot_and_gsc_sync

Multi-CMS SEO/GEO Autopilot, GSC Heartbeat & Shared Action Board

Action

Closed-loop SEO & GEO Autopilot, Google Search Console (GSC) rank decay heartbeat, Multi-CMS staging (WordPress, GitHub PR for Next.js/Astro, Webflow, Ghost, Shopify, custom webhook), and Shared Action Board orchestrator (actions grouped by type and sorted by expected_clicks descending, assigned to Chief of SEO, Analyst, Writer, or Refresher, where night routines always stop at "Ready for review" for human approval).

inputSchema

{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "action",
    "domain"
  ],
  "properties": {
    "action": {
      "type": "string",
      "enum": [
        "save_project_context",
        "track_keywords_rank",
        "gsc_performance_sync",
        "stage_rewrite_for_approval",
        "approve_and_publish",
        "rollback_change",
        "list_action_board",
        "update_action_board_item"
      ],
      "description": "Autopilot operation to execute: manage project memory, run weekly rank decay checks, stage/approve/rollback CMS rewrites, or list/update the Shared Action Board."
    },
    "domain": {
      "type": "string",
      "description": "Target project domain (e.g., \"argosor.com\")."
    },
    "brandName": {
      "type": "string",
      "description": "Optional brand display name for project memory and Answer-First rewrite staging."
    },
    "brandVoiceGuidelines": {
      "type": "string",
      "description": "Optional brand voice and tone guidelines stored in persistent project memory."
    },
    "cmsPlatform": {
      "type": "string",
      "enum": [
        "wordpress",
        "github_pr_nextjs_astro",
        "webflow",
        "ghost",
        "shopify",
        "custom_webhook"
      ],
      "description": "Target CMS or Git deployment target for staged rewrites and 1-click rollbacks. Defaults to \"github_pr_nextjs_astro\"."
    },
    "keywords": {
      "type": "array",
      "items": {
        "type": "string",
        "description": "Target keyword phrase to track."
      },
      "description": "Optional list of target keywords to save or check for 3+ position decay."
    },
    "targetUrl": {
      "type": "string",
      "description": "Target page URL when staging an Answer-First SEO/GEO rewrite for human approval."
    },
    "targetKeyword": {
      "type": "string",
      "description": "Primary keyword when staging a page rewrite."
    },
    "changeId": {
      "type": "string",
      "description": "Staged change ID when calling approve_and_publish or rollback_change."
    },
    "actionItemId": {
      "type": "string",
      "description": "Action Board item ID when calling update_action_board_item."
    },
    "owner": {
      "type": "string",
      "enum": [
        "Chief of SEO",
        "Analyst",
        "Writer",
        "Refresher"
      ],
      "description": "Specialist agent owner on the Shared Action Board."
    },
    "status": {
      "type": "string",
      "enum": [
        "Todo",
        "Working",
        "Blocked",
        "Ready for review",
        "Done"
      ],
      "description": "Workflow status on the Shared Action Board. Night routines automatically cap status at \"Ready for review\"."
    },
    "expectedClicks": {
      "type": "integer",
      "minimum": 0,
      "maximum": 1000000,
      "description": "Estimated monthly organic + AI click gain used to sort the Shared Action Board descending."
    },
    "nightRoutine": {
      "type": "boolean",
      "description": "When true, enforces the night routine guardrail so items stop at \"Ready for review\" (requires_human_approval: true)."
    }
  }
}

Discovery & Markdown

Every HTML page sends an RFC 9727 Link header pointing agents to the catalogs below, and every page has a Markdown twin: append .md or send Accept: text/markdown.

/.well-known/mcp.json	MCP discovery manifest
/.well-known/mcp/server-card.json	MCP server card
/.well-known/api-catalog	RFC 9727 API catalog (linkset)
/.well-known/ai-catalog.json	Agent resource catalog
/.well-known/agent-skills/index.json	Downloadable SKILL.md files with SHA-256 digests
/.well-known/glama.json	Glama connector manifest
/.well-known/agent.json	A2A agent card
/openapi.json	OpenAPI 3.1 specification
/llms.txt	LLM-friendly overview (plus /llms-full.txt)