Content Intelligence

Question 1

“Can we keep the agents to read-only access?”

Answer

Partly answered

The site covers choosing which records each agent can read, which actions it takes alone and which need approval, with read, draft and send rules per agent. It never states plainly that an agent can be limited to read-only, with no drafting, logging or record changes, or what each agent can still do in that mode.

Section on the site	Status	Ticket
/products/agentsDecide what each agent can do (card)	Partly answers it	State whether agents can be set to read-only on /products/agents Ticket Paste this into your ticketing tool. State whether agents can be set to read-only on /products/agents Who needs it: CTOs and IT or security reviewers checking a vendor Section: /products/agents › Decide what each agent can do (card) What is missing: - No explicit statement that an agent can be limited to read-only access with no write actions. - No explanation of what a read-only agent can still do, such as suggesting, drafting or logging activity on records. - No note on whether automatic actions like the Prospecting Agent's activity logging can be switched off. Where: Update this section: add a fourth line under "Decide what each agent can do", after "Rules apply as soon as you save them". Related sections: Everything you need to run agents responsibly · Agents that ask. People who decide. What to do: 1. Confirm with product whether each agent can be set to read-only and which automatic writes, such as activity logging, a read-only setting stops. 2. Add a line to the card naming the read-only option and what it allows. 3. Link the line to the Permissions and approvals card for the full set of controls. Done when: /products/agents states whether an agent can be kept read-only and what it can still do in that mode. From: conversation ca452194d968, question 1: "Can we keep the agents to read-only access?" Who needs it: CTOs and IT or security reviewers checking a vendor What is missing No explicit statement that an agent can be limited to read-only access with no write actions. No explanation of what a read-only agent can still do, such as suggesting, drafting or logging activity on records. No note on whether automatic actions like the Prospecting Agent's activity logging can be switched off. Where Update this section: add a fourth line under "Decide what each agent can do", after "Rules apply as soon as you save them". Related sections Everything you need to run agents responsibly · Agents that ask. People who decide. What to do Confirm with product whether each agent can be set to read-only and which automatic writes, such as activity logging, a read-only setting stops. Add a line to the card naming the read-only option and what it allows. Link the line to the Permissions and approvals card for the full set of controls. Done when: /products/agents states whether an agent can be kept read-only and what it can still do in that mode.

Section on the site

Status

Ticket

/products/agentsDecide what each agent can do (card)

Partly answers it

State whether agents can be set to read-only on /products/agents

Who needs it: CTOs and IT or security reviewers checking a vendor

What is missing

No explicit statement that an agent can be limited to read-only access with no write actions.
No explanation of what a read-only agent can still do, such as suggesting, drafting or logging activity on records.
No note on whether automatic actions like the Prospecting Agent's activity logging can be switched off.

Where

Update this section: add a fourth line under "Decide what each agent can do", after "Rules apply as soon as you save them".

Related sections: Everything you need to run agents responsibly · Agents that ask. People who decide.

What to do

Confirm with product whether each agent can be set to read-only and which automatic writes, such as activity logging, a read-only setting stops.
Add a line to the card naming the read-only option and what it allows.
Link the line to the Permissions and approvals card for the full set of controls.

Done when: /products/agents states whether an agent can be kept read-only and what it can still do in that mode.

Answer full page

Yes, you stay in control. In Agent Hub you decide what each agent can read, what it can do on its own and which actions wait for a person, and you can review every action. The docs below describe the four scope levels you can set for an agent.

Open the answer page ↗

Sections in the answer page	Status
/products/agentsPermissions and approvals (card)	Answers it
/products/agentsAgents that ask. People who decide.	Answers it
/solutions/govern-aiFour controls for every agent	Answers it
/products/agentsDecide what each agent can do (card)	Partly answers it

Questions and answers are shown as recorded. Status, what is missing and what to do are our analysis of the live site, read 22 Oct 2026.

Conversation