Temei Library
Governed AI agents, written down
Open documents from a UK consultancy that builds Copilot Studio agents for itself first: how to make them work, and the rules that keep them safe.
- 6 KB articles
- 6 governance documents
- Free to read and adapt
Knowledge base
KB-001
Why DLP blocks your Copilot Studio agents, and how to stop it happening
An agent that works in Preview and then fails the moment it needs a trigger is rarely a bug. It is two policy systems disagreeing about one connector. Here is how to read them, fix them and keep them aligned.
KB-002
Making an agent run on its own: harness first, trigger second
The trigger you need depends on which runtime powers the agent. Choose the wrong door and Power Automate shows you an empty list of agents, with no error to explain why.
KB-003
Giving development its own rules without loosening production
Policies in Power Platform apply to whole groups of environments. If development shares a group with production, every exception you make for a builder is an exception in production too. Here is how to separate them cleanly.
KB-004
Letting an agent write to SharePoint without letting it write anything
Instructions tell an agent what it should do. Only the tools you give it decide what it can do. For agents that update business records, put the rules in the tool, not in the prompt.
KB-005
Should your agent live in Copilot chat, run on its own, or both?
Publishing an agent to Microsoft 365 Copilot and Teams puts it where people already work. Making it autonomous takes people out of the loop. They solve different problems, and the choice decides cost, risk and who is accountable.
KB-006
The governance pyramid for Copilot Studio agents
Six layers, from a one-page charter down to the records that prove it, with the Microsoft control that enforces each one. Built by a small consultancy that runs governed agents for itself first and for clients next.
Governance documents
The charter, policies, standard and procedures Temei works to, published in full as working drafts. Read them, adapt them, use them.
CHR-00
CHR-00 Temei AI Charter
Why Temei uses AI agents, the eight promises we keep, and the red lines no agent crosses at any level.
POL-01
POL-01 Agent Autonomy Policy
How much an agent may do on its own, how it earns more, and how it loses it. The ENGRAM Trust Pyramid applied to Copilot Studio.
POL-02
POL-02 Agent Data and Privacy Policy
Which data an agent may touch at each trust level, where it may go, how long anything is kept, and who answers for it under UK GDPR.
PRO-02
PRO-02 Agent Promotion Procedure
How an agent moves up one trust level at a time, with an evidence pack, a cooling-off review and a signed Release Certificate.
PRO-05
PRO-05 Agent Incident, Suspension and Retirement Procedure
Stop first, understand second: three switches to stop an agent, severity rules, incident steps, and how to retire an agent completely.
STD-01
STD-01 Agent Build Standard
The exact settings every agent must have before it is called built: environment, naming, instructions, tools, guarded writes, triggers and six tests.