What This Looks Like
Multiple workflows, agents, checks, automations, or downstream paths depend on the same tool, connector, API, MCP server, or integration. When that tool changes, slows down, fails, or becomes unavailable, many separate workflows are affected at once.
Why Users Blame AI
The failure may appear as many AI tasks failing at the same time. Users may think the model, agent, or prompt is broken, when the actual issue is a shared tool dependency that sits underneath multiple workflows.
What to Check First
- Which workflows depend on the same tool or connector
- Whether the tool is a single point of failure
- Whether fallback behavior exists when the tool is unavailable
- Whether tool schema or permission changes affect multiple workflows
- Whether review queues or downstream steps depend on the tool output
- Whether each workflow truly needs direct access to the same tool
When This Is AI-Adjacent
Use this AI-Adjacent Issue when many workflows depend on one shared tool or integration. If the tool can act without responsible authority, use the related Workbench Issue for tool authority. If a small tool change causes large downstream effects, use the related propagation Issue.