.svg)
Coding-agent session evidence and workstation monitoring.
Investigate agent sessions, credential exposure and changes to local agent configuration. Follow a finding to the workstation and the recorded activity behind it.
Session findings linked to recorded activity.
Inspect prompts, tool calls and responses in sequence. Trace credential exposure through recorded session evidence instead of investigating an isolated alert.
Read services/payments/.envAgent configuration changes and risk assessments.
Compare workstation snapshots to find new or modified skills, hooks and MCP configuration. Open the file assessment to understand the risk.
Session and configuration collection coverage.
Session sources include Claude Code, Codex, OpenCode and supported Cursor and VS Code chat sources. Configuration inventory covers agents, skills, hooks, MCP servers and editor extensions separately.
Workstation reporting status and collection gaps.
Review reporting state, sensor versions and collection health. Identify stale or missing evidence before relying on a fleet-wide assessment.
Raw-evidence access and retention.
Set raw-evidence access separately from indexed session metadata. Raw transcripts and collected files can contain sensitive data; inventory redaction does not sanitize those artifacts.
Agent enforcement and session investigation in one platform.
Use Agent Hooks for guidance and supported enforcement inside the coding agent. Use the Sensor for session and workstation evidence.
Deployment through device management.
Deploy to macOS and Windows workstations. Enroll each machine with a deployment credential, then use its machine identity for subsequent uploads.
Investigate agent behavior using recorded evidence.
Connect workstation configuration, session activity and repository context in Coding Agent Security.
