Policies


URL	`/policies/agentic-use`
Sidebar	Policies → Agentic Use
Roles	Admin · Analytics Viewer, reading

Agentic Use is where you decide what your coding agents are allowed to do — which shell commands they may run and which MCP tools they may call. Policies in the sidebar expands to its groups; Agentic Use is the first. Opening /policies directly lands here too.

What’s on the page

Tab strip

Tab	URL	Contents
Tool Policies	`/policies/agentic-use`	The policy list
Policy Health	`/policies/policy-health`	How policies are performing, with Overview and Recently Edited. Shown only where your organization has it turned on, and only to an Admin — an Analytics Viewer who opens the URL is sent to Analytics
Environment Targets	`/policies/environment-targets`	The hosts, databases, cloud resources, secrets, repositories, paths and other targets your agents have actually touched, each tagged by sensitivity. Shown only where your organization has it turned on. An Analytics Viewer can read it but cannot set a tag

Policy list

Column	Contents
Name	The policy name
Type	Terminal command or MCP
Command Family	The family it matches
Target Pattern	The pattern being matched
Action	What happens on a match
User Groups	Which groups it is scoped to, or org-wide
Status	Whether it is enabled

Environment Targets

Every host, database, cloud resource, secret name, container image, repository, branch, file path and other target your agents have touched. Each row carries the command family it came from, how many developers used it, when it was last seen, how many of those uses were high risk, and the total number of times it has been seen. You tag each one Low, Med or High sensitivity. From then on, commands against that target are scored in the band you named rather than the one Unbound would have guessed — your own knowledge of which database is production wins. Commands already classified keep the score they were given. Untagged paths fall back to defaults: /etc, /usr, /var, /boot, /sys, ~/.ssh, ~/.aws, ~/.gnupg and .env files read as High, and ~/.config, .github, Dockerfile and docker-compose* as Med. Filter by command family or by sensitivity (including Untagged, which is the working list), and search for a target by name.

Filters and controls

Search and filter the policy list by name, type and status. Group scoping is set per policy, drawn from Users → User Groups.

What you can do

Action	Result
Create Policy	Opens the type picker, below
Preview Impact	In the policy form, before you save: shows what the rule you have entered would have matched over the last 7 days. Always offered for terminal policies; for MCP policies the button appears only where your organization has it turned on
Policy Packs	Applies a curated set of starter policies in one step. The button carries a count — `applied/available` once you have applied at least one pack, otherwise the number of packs available. Shown only where your organization has it turned on, and not to an Analytics Viewer
Logs on a row	Opens the Analytics — Agentic Use page filtered to that policy’s matches
Select a policy	Opens its own page at `/policies/agentic-use/[id]`, described below
Select a match on a policy page	Opens the match panel, below
Delete a policy	Takes it off the list and out of enforcement — it stops matching, and the dashboard offers no way to bring it back. Everything it already matched stays in Analytics, so deleting a rule does not cost you its history. A policy that came from a policy pack cannot be deleted this way; remove its pack instead

A policy’s own page

Section	Contents
Identity strip	The policy’s name, type and scope
Insights	A KPI row and breakdowns showing how often it has matched
Matches	The runs it caught, with a filter bar to narrow them

Match panel

Field	Contents
Timestamp	When the run happened
AI tool	The tool that ran it
Action	What the policy did
Repository	The repository in play
Working directory	Where the command ran

The policy types

Type	What it does
Terminal Commands	Control which shell commands AI coding tools can execute
MCP Actions	Control which MCP tools AI coding agents can use
MCP Probe	Allow an MCP server’s tools only when an identity probe passes a check · carries a Beta badge
MCP Identity Signals	Flag MCP tool-call return values that reference an identity outside your allowlist · carries a Beta badge
Skills	Require AI coding agents to invoke specific skills when policies match

Terminal Commands and MCP Actions are always in the picker. MCP Probe, MCP Identity Signals and Skills each appear only where your organization has that one turned on. A skill policy opens at /policies/skill-policies/[id]. That detail page is turned on separately from skill policies themselves; where it is not on, selecting a skill policy opens its edit form instead — and an Analytics Viewer, who cannot open that form, gets nothing. Probe policies are created from Create Policy → MCP Probe.

Ask AI

/policies/agentic-use/ask-ai drafts a policy from a plain-language description for you to review.

Messages you may see

Message	What it means
`Failed to save policy`	The policy was not saved. Your entries remain on screen, so saving again is enough
`Failed to delete policy. Please try again.`	The policy is still in place
`Failed to create policies.`	A bulk creation did not complete. Check the list before retrying
`Could not load discovered skills.`	The skill picker could not list the skills found on your fleet, so a skill policy cannot be built from it yet
`Could not load this skill's body.`	The picker could not show what that skill tells the agent to do
`Failed to load policy matches`	The match history could not be fetched. The policy continues to run
`Failed to preview impact`	The preview could not be generated. The policy can still be saved
`Failed to apply pack` / `Failed to remove pack`	A policy pack did not finish. Check the list before retrying
`Failed to save probe policy`	The probe policy was not saved
`Could not load probe policies.`	The probe list could not be fetched
`Could not load tools.`	The MCP tool picker could not load its options
`Failed to load environment targets. Please try again.`	The Environment Targets tab could not load
`Failed to load user groups. Refresh the page to retry.`	Group scoping options could not be fetched
`Failed to create label` / `Failed to delete label`	A policy label was not saved or removed

Page	For
Analytics — Agentic Use	What these policies matched
Policies — Data	Guardrails for content
Policies — Usage	Budgets, repository scope and account access

Getting Started

CLI

Analytics

Hooks API

Integrations

MDM Integrations

Connections

Support & Troubleshooting

Policies — Agentic Use

What’s on the page

Tab strip

Policy list

Environment Targets

Filters and controls

What you can do

A policy’s own page

Match panel

The policy types

Ask AI

Messages you may see

​What’s on the page

​Tab strip

​Policy list

​Environment Targets

​Filters and controls

​What you can do

​A policy’s own page

​Match panel

​The policy types

​Ask AI

​Messages you may see

​Related

What’s on the page

Tab strip

Policy list

Environment Targets

Filters and controls

What you can do

A policy’s own page

Match panel

The policy types

Ask AI

Messages you may see

Related