Policies
Policies
Set the approval strength OSuite should enforce globally or for one agent.
Runtime policy scope
Choose what this policy edit applies to.
Workspace default applies everywhere. Agent-specific policy overrides the default for one connected runtime, such as ChatGPT Apps, Codex, or Claude Code.
Workspace default
This is the fallback policy for every runtime.
Policy applies to
Workspace coverage
Applies to 0 connected agents. No agent-specific overrides are changed.
0 agents0 agent guard rules
Package boundary
Package depth controls seats, governed volume, and replay-ready evidence while policy logic stays here.
Keep profiles, packs, and raw rules here. Use billing only when the workspace needs more volume, more operators, or deeper evidence exports.
Current plan: FreeFree until 2 Aug 2026
Approval strength
Tune the workspace default policy.
Choose a baseline, adjust the five sliders, then save. OSuite will apply the saved profile before connector defaults on future actions.
Balanced
Start here for most teams. External sharing is reviewed, destructive actions are gated, internal work stays fast.
Shadow mode: review high riskBest for: General enterprise rolloutPack: SMB Safe
Profile posture
Use allow, approval, or block for each boundary. The saved profile applies to the selected scope above.
Current lens: Balanced
Core posture
Managed from Policies
Policy-owned decision weighting
PCAA stays stable here. This lens controls how the workspace orders and escalates decisions after actions have already been turned into governed objects.
Current lens: Balanced
View scoring details
Intent & executionData & privacyImpact & reversibilityTrust & adversarial pressureGovernance fit
Approval gates
External sharing requires approvalPublic links require approvalProduction changes require approvalIdentity admin escalation requires approval
Monitoring posture
Review high-risk actions while allowing routine internal work
Scoring lens: Balanced
This becomes the default for all agents without their own override.
0
Total Policies
0
Active
0
Blocks (24h)
0
Evaluations (24h)
Decision impact
Use Decisions for history and replay. Stay here for policy authoring and simulation.
Approval crossings
0
Recent governed actions that crossed a human checkpoint.
Fleet scope
0
Agent boundaries currently addressable by policy.
Evaluated actions
0
Decisions evaluated across the last 24 hours.
Guard Policies
Loading policies...
Policy Test Runner
Click "Run Tests" to validate all policies against their test cases.
Proof Report
Generate a proof report to document policy compliance status.