VISIBILITY
See your coding-agent footprint
Identify supported agents, skills, MCP servers, plugins, and configuration files on managed devices. Review the files and devices behind a finding.
Coding agent security
Keep coding agents in bounds, from skills to live actions.
See the agents and skills on managed devices, investigate what they do in live sessions, and stop covered actions that violate policy.
Visibility, monitoring, enforcement
Know what agents can access. Control what they do.
Workstation Lens connects agent-file findings to session activity and policy decisions on supported managed devices.
MONITORING
Understand what happens in a session
Follow supported prompts, tool calls, commands, file access, and outbound connections to investigate unexpected behavior.
ENFORCEMENT
Stop actions that violate policy
Block covered commands, tool calls, resource access, or destinations before execution under a configured policy.
Agent and skill visibility
See the skills and connections behind your coding agents.
Identify supported agents on managed devices. Inspect their skills, MCP servers, plugins, and configuration files, then locate the devices affected by a finding.
SkillsMCP serversPluginsRulesHooksConfiguration files
Agent inventory / findings
Live session / policy eventBlocked
Runtime control
Put policy between the agent and a risky action.
Investigate tool calls with session context
Review prompts, commands, file and credential access, and outbound connections in supported sessions.
Block covered actions before they execute
Use configured policies to stop prohibited tool calls, commands, data access, or destinations at supported enforcement points.
Investigation and response
Trace a finding from file to fleet.
Review file findings alongside live activity and policy outcomes. See where the same risky file appears so the response reaches affected managed devices.
Flagged skill
The file and risky instruction
Session activity
What the agent attempted
Policy outcome
Observed or confirmed blocked
Affected devices
Where the same file appears
Frequently asked questions
What to know before deployment.
Coverage, enforcement, and how to verify a policy on your managed devices.
Which coding agents can Workstation Lens discover?
Workstation Lens identifies supported coding agents, including Claude Code, Cursor, Codex, and GitHub Copilot, on managed devices. Scanning, live-session visibility, and blocking coverage depend on the agent and deployment.
What does Repello scan on a developer's device?
Workstation Lens inspects supported skills, MCP servers, plugins, rules, hooks, and configuration files for risky instructions, permissions, commands, and destinations. Each finding points to the relevant file and managed device so your team can investigate it.
Which live coding-agent actions can security investigate?
In supported sessions, security teams can review prompts, tool calls, commands, file and credential access, outbound destinations, and policy decisions. The session context shows what the agent attempted and whether a policy acted; it does not expose the model's private reasoning.
Can Repello block a prohibited action without disabling the agent?
Yes. Where runtime enforcement is supported, Workstation Lens can block a prohibited tool call, command, resource access, or destination while allowing permitted work to continue. Blocking requires an active policy; discovering an agent alone does not enable it.
How do we verify a policy actually blocked an action?
Check the session record for the attempted action and its policy outcome. A confirmed block shows the action was stopped at a covered enforcement point; a scan finding or monitoring alert shows detection, not prevention.
How is this different from code scanning or EDR?
Code scanners assess the software being built, while EDR monitors endpoint activity. Workstation Lens focuses on the coding agent: its connected skills and MCP servers, live actions, and policy decisions on managed devices. It complements both rather than replacing them.
See coding agent security in action
Get a personalized Repello demo.
See a skill scan, a live coding-agent session, and a policy block together. Review the evidence and identify the managed devices that need attention.