Workbench / AI-Adjacent Issues

One Workspace Affects Another

State, memory, files, permissions, settings, tools, or workflow effects from one workspace affect another workspace where they should not apply.

What This Looks Like

A setting, memory, file, permission, tool result, workflow state, or local rule from one workspace affects behavior in another workspace. The user may see context bleed, unexpected formatting, wrong assumptions, shared state changes, or actions appearing outside the workspace where the task began.

Why Users Blame AI

The user experiences the AI as mixing projects, accounts, contexts, or tasks. The underlying cause may be workspace configuration, shared memory, connector scope, permissions, file access, agent state, or product behavior that does not isolate workspaces the way the user expects.

What to Check First

When This Is AI-Adjacent

Use this AI-Adjacent Issue when workspace isolation may be failing or unclear. If context crosses between tasks, use the related Workbench Issue for context leakage. If an agent actually modifies unrelated state, use the related Issue for agent-modified unrelated state.

Common Ways People Describe This

Metadata

Code
ADJ-0008
Layer
ai_adj_issue
Version
0.1.0
Updated
2026-05-10T00:00:00Z