adlyse-workflows
Endpoint: https://mcp.adlyse.com/mcp/workflows/
Build and run Adlyse’s operational workflows from outside the platform. This surface owns node-based workflow graphs, AI colleagues and handoffs, reusable skills, schedules, data tables, and institutional-memory files.
Use this server to integrate Adlyse workflows into an external orchestration: e.g. invoke generate_lead_report nightly from a scheduler, or enrich_contact from a CRM webhook.
To have Claude author workflows or colleagues against this server, install the
exported workflow_builder or agent_builder skill — see
Install the builder skills.
Tools#
| Tool | Purpose |
|---|---|
list_workflows | Every root workflow in your org |
get_workflow | Detail including input / output schema |
run_workflow | Kick off an async run; returns a run_id |
get_run_status | Poll status, output, and error |
run_assistant | Run an agent once, now, with a message (test a colleague) |
get_assistant_run | Status, reply, tool calls and cards of one headless agent run |
update_table / delete_table | Rename or delete one of your data tables |
update_table_column / delete_table_column / reorder_table_columns | Edit a table’s schema |
The full list — workflow graphs, agents, handoffs, skills, memory, releases,
schedules, and data tables — is in the sidebar. The same operational tools are
included in the combined adlyse-all endpoint.
Common errors#
error_type | Typical cause |
|---|---|
auth_error | Missing / invalid API key |
not_found | Workflow id / run id not in your org |
invalid_input | Input failed the workflow’s trigger-schema validation |
execution_error | Run couldn’t be scheduled (rare — usually transient) |