Productivity agents
Productivity agents summarize documents, manage calendars, search email, create tasks, and share files across multiple systems. Their usefulness comes from broad connectivity, which also increases their potential blast radius.
Tests malicious documents, email-based prompt injection, cross-application leakage, unsafe sharing, destructive actions, and poisoned MCP responses.
Checks every action against the user's request and requires approval for destructive or sensitive operations.
Productivity agents summarize documents, manage calendars, search email, create tasks, and share files across multiple systems. Their usefulness comes from broad connectivity, which also increases their potential blast radius.
Tests malicious documents, email-based prompt injection, cross-application leakage, unsafe sharing, destructive actions, and poisoned MCP responses.
Checks every action against the user's request and requires approval for destructive or sensitive operations.
A user asks the agent to summarize a roadmap and schedule a review. Hidden instructions inside the document ask it to share the finance folder with an external address. dFence permits the summary and calendar action but blocks the unrelated file share.
For every agent, dwaar asks four questions: What is the agent supposed to do? Which data and tools should it reach? How could that workflow be manipulated? What should happen when its behavior drifts?
Document the happy path, permitted tools, data boundaries, and approval points.
dForge attacks the workflow before launch using context-specific prompt, tool-use, and multi-turn tests.
dFence evaluates runtime intent before sensitive tool and MCP calls, then allows, blocks, constrains, redacts, or requests approval.
Turn verified findings and runtime events into stronger policies and repeatable security tests.
Choose a use case to see its risks, adversarial tests, runtime controls, and security decisions in context.
dwaar combines agent-specific adversarial testing with intent-aware runtime enforcement, helping teams deploy capable agents without relying on static scopes alone.
