MCP Servers
Adlyse exposes its performance-marketing data and judgments over the Model Context Protocol so external AI clients — Claude Code, Claude.ai, ChatGPT Desktop, or your own orchestration — can read from and write into Adlyse the same way our in-app assistant does.
Five scoped MCP servers (plus an aggregate convenience endpoint) run behind a
single combined host. All are scoped to your organization by your API key.
adlyse-admin is intentionally separate because it exposes organization
hierarchy information; operational authoring is part of workflows and the
aggregate endpoint.
The servers#
| Server | Path | What it exposes |
|---|---|---|
adlyse-insights | /mcp/insights/ | ABI, waste, safe-budget, impact analysis, suggestions, competitor intel (Brand Watch), strategy, guidelines |
adlyse-reports | /mcp/reports/ | List / generate / fetch performance reports with presigned PDF URLs |
adlyse-ads | /mcp/ads/ | Cross-platform campaign & ad reads, creative library, experiments; gated writes |
adlyse-workflows | /mcp/workflows/ | Workflows, agents, handoffs, skills, schedules, tables, and institutional memory |
adlyse-admin | /mcp/admin/ | Organization details and direct sub-organization enumeration |
adlyse-all | /mcp/all/ | Aggregate of insights + ads + reports + workflows under one connection (admin excluded) |
- Production host:
https://mcp.adlyse.com - Local dev host:
http://localhost:8766(run./start-mcp.shfromserver/)
Getting started#
- Authentication — how API keys and OAuth work
- Quick start — wire Claude Code (or any MCP client) to Adlyse in under a minute, including the one-command
npx adlyse-mcpinstaller
Conventions#
Every tool response carries a source tag — e.g. adlyse.dashboard_compute@v1 — so callers can verify which computation produced the data and reason about version drift.
Writes are preview-safe. Platform write tools (campaign/ad-group/ad status and campaign budget on adlyse-ads) accept a preview: true parameter that returns the intended change without calling the platform API. Apply with preview: false. Sensitive writes (experiment control) additionally require a user-bound credential — OAuth or a personal API key.
Writes auto-log. On a successful mutation, Adlyse writes a DecisionLogEntry with created_via="tool" and a short fingerprint of the calling API key. This feeds back into get_impact_analysis so future metric shifts correlate with past decisions. See Home for how those decisions surface and get measured.
Admin is explicit. adlyse-admin is not part of the default aggregate
connector. OAuth tokens need the admin scope and users need agency-management
permission. Organization API keys continue to authenticate organization-scoped
admin access.
Errors are structured. Failed tool calls return { "error": "...", "error_type": "..." } rather than raising. Common error_type values: auth_error, not_found, validation_error, unsupported_for_platform, platform_error.