Information Technology · infrastructure-and-cloud Infrastructure & Cloud
Cost Optimization Agent
task agent in the workflow layer, owned by Tomas Berger (Principal Engineer, Agent Platform). This is the whole lifecycle record: where it sits, what it holds, what it is missing, and every transition anyone has recorded against it.
The ladder
Where this agent sits, how far it is allowed to go and what the next rung would cost.
Nothing on file. This agent was already running when the gate went live on Mar 3, 2026.
- ·Independent evaluation review by someone outside the team that built the agent
- ·Ninety days at delegated with the full override record intact
- ·Kill switch tested on this agent, with a date and a measured stop time
- ·Named owner for every exception class this agent can raise
- ·Executive sign-off recorded against the accountable function leader
This agent was already running at this stage when the lifecycle gate was introduced. Its entry evidence was never produced. It is frozen at its current stage — it cannot be promoted, and the missing evidence is listed against it — but it was not demoted, because demoting a working line to satisfy a records gap would have been theatre.
Any single priority-one exception attributed to the agent, an evaluation regression beyond the tolerance band, or a service-level breach attributed twice in a quarter returns it to delegated automatically.
Wired to a live threshold. It fires and moves the agent down without anyone deciding to.
Running at a stage it reached before the gate existed. The platform holds no promotion record for it.
Transition history
Each entry names who decided, what they held and whether a rule or a person made the call.
No transitions recorded. This agent predates the gate, which went live on Mar 3, 2026. It reached its current stage before anyone was keeping a record, and that absence is why it is frozen rather than shown as compliant.
Evaluation record
The golden set this agent is measured against, and every run scored against it.
Nothing has been written that this agent can be tested against. Without a golden set there is no evidence a promotion could cite, which is exactly why this agent cannot move up regardless of how well it appears to be performing.
Every evaluation on this page ran against golden sets we wrote ourselves, on a modeled estate. A passing suite proves an agent behaves the way we specified, not that the specification is right, and no evaluation here has been reviewed by anyone outside the team that built the agent.
This agent has no golden set, so there is nothing here to score. The record shows an absence rather than a pass, which is the honest reading: an untested agent is not a safe agent that has not been checked yet, it is an agent whose behavior nobody has established.