run proof
Run logs
Capture agent, task, prompt source, command evidence, verification result, and handoff status.
Codex / Claude Code / Cursor ops templates
AI agent ops template pack for proof-first teams.
A single map for run logs, verification ledgers, cost tracking, ROI, procurement, security review, acceptance tests, prompt regression, model routing, workflow comparison, and client delivery proof.
Template map
verification
Verification ledger
Track claims, commands, artifacts, screenshots, reviewer notes, and unresolved proof gaps.
finance
Cost and ROI
Connect AI-agent cost tracking to cleanup hours avoided, delivery confidence, and team-license approval.
approval
Procurement and invoice
Give managers the purchase packet, receipt language, invoice request copy, and approval summary.
risk
Security review
State access boundaries, paid-ZIP delivery limits, audit evidence, and post-payment handling.
quality
Acceptance and prompt regression
Keep completion, repeated prompts, model changes, and scope drift testable before handoff.
routing
Model routing
Choose the right model, review path, escalation rule, and cost guard for each run type.
workflow
Workflow comparison
Compare chat history, generic Notion, spreadsheets, PM tools, and the full Agent Ops pack.
delivery
Client delivery
Package agency or consulting work with run proof, delivery acceptance, and handoff language.
Request packet
A request is not revenue. Count revenue only after checkout, receipt, payout, or seller-dashboard evidence tied to the paid template pack.