Operations loops

Recurring upkeep: status reports, queue triage, error sweeps, cost watching and agent health checks.

Loops › Operations

Timed, small, verified repo improvements: one fix per cycle, one commit each.

Claude Code, Codex, Cursor and other coding agents · Stops when: the time box ends or nothing small is left

Updates STATUS.md each week from your real sources and carries open threads forward.

Any agent with your tools connected · Stops when: STATUS.md is current; it runs again next week

Works through a queue (issues, tickets, inbox) one item at a time and logs a note for each.

Any agent · Stops when: the queue is empty or the batch size is reached

Reads production errors, groups them, and fixes the most frequent real one.

Claude Code, Codex, Cursor and other coding agents · Stops when: the top errors are fixed or ticketed

Checks whether source-of-truth docs have drifted from the code and flags contradictions in one bounded pass.

Claude Code, Codex, Cursor and other coding agents · Stops when: one pass is complete and the drift list is written

Turns API spend and provider failures into a short list of decisions, before the bill surprises you.

Any agent · Stops when: the decision list is written; runs again on schedule

Checks that your agents are reachable, stay in their roles and have no stuck jobs, and writes a receipt.

Multi-agent setups · Stops when: the receipt is written

Checks whether your autonomous system is improving, going in circles, wasting tokens or losing reliability.

Claude Code, Codex, Cursor and other coding agents · Stops when: the report is written