copilot-credit-spike-contain

Stop a Copilot Studio agent burning credits faster than it was budgeted for, before deciding whether the consumption is genuine.

Procedure

  1. start start. Goes next to s1
  2. check Confirm the spike in the consumption grid. Goes on_fail to __end_aborted; next to s2
  3. end aborted. Terminal
  4. check Check for an existing per-agent limit. Goes on_fail to s9; next to s3
  5. action Apply or lower the per-agent limit, RW (reversible-write). Goes on_fail to s9; next to s4
  6. decision Is consumption still rising?. Goes if rising to s5; if default to s6
  7. action Disable the agent in its environment, DEL (destructive), requires approval. Goes rollback to s7; on_fail to s9; next to s6
  8. action Re-enable the agent, RW (reversible-write). Goes next to __end_success
  9. check Determine the cause before re-enabling. Goes on_fail to s9; next to s8
  10. escalate Escalate to the Power Platform admin. Terminal
  11. action Record the incident and its resolution, RW (reversible-write). Goes next to __end_success
  12. end success. Terminal

Open the graph on its own page — larger, printable, and an address you can send.

Walk it without running it — you answer what your environment would answer, and the path, the capabilities, the approvals and the distance back to an undo come out of the document. Nothing executes; there is no Run button here and there will not be one.

Trust

T2 · attested 2026-10-11 · runbooks.directory

Nobody has attested this record. That is the ordinary case: T2 is what the catalog can check on its own, and T3 needs somebody to run the procedure and sign for it.

  • Link liveness has not been checked: it needs a network and the catalog builds without one. Run pnpm gates --online where there is one.

Steps

  1. s1 check — Confirm the spike in the consumption grid
  2. s2 check — Check for an existing per-agent limit
  3. s3 action — Apply or lower the per-agent limit (reversible-write)
  4. s4 decision — Is consumption still rising?
  5. s5 action — Disable the agent in its environment (destructive) — requires approval
  6. s6 check — Determine the cause before re-enabling
  7. s7 action — Re-enable the agent (reversible-write)
  8. s8 action — Record the incident and its resolution (reversible-write)
  9. s9 escalate — Escalate to the Power Platform admin

Capabilities

This is a privilege request. Check it against your own allowlist before running.

    Provenance

    Source
    https://github.com/microsoft/ai-agent-runbooks/blob/083f33eec889/02-patterns/Copilot-Credits-Cost-Control/Copilot-Credits-Cost-Control-Runbook.md
    Fetched
    2026-09-04
    Source last checked
    not since it was captured on 2026-09-04 — recorded as current
    Licence
    MIT
    How it got here
    Rewritten here from the source above, under its licence. The words and the steps are this catalog's; the procedure it describes is not this catalog's invention, and the source is where to check it.
    Highest risk in this procedure
    destructive

    Take it with you

    Run it under supervision

    This procedure contains destructive steps. Run it here, where the supervisor is the thing invoking the commands and therefore sees every one of them: each step is proposed before it happens and a gate stops for you. Nothing about the run leaves this machine.

    # destructive steps: every one stops for you before it runs.
npx @runbooks/cli run adapted/copilot-credit-spike-contain@1.0.0 --as "$(whoami)" \
  --input environment=<the Power Platform environment the spike is in> \
  --input agent=<the agent named in the consumption grid>