Operator Agent Skills

Agentic Back Office

Source-grounded pack with Audit an Agentic Back Office, Build a Decision Evidence Packet, and Design a Back-Office Loop, plus 2 more.

Use this pack when

Choose this pack when the work touches Agentic Back Office and you need a bounded workflow with explicit inputs, a defined output, and human review gates.

Included workflows

Choose the job in front of you.

5 source-grounded skills. Load one workflow at a time.

Audit an Agentic Back Office

Use this workflow when assessing whether a back-office workflow is ready to launch, widen, or deepen agent delegation by testing control integrity, event replay, reversibility, exception handling, monitoring, and accountable ownership.

Output Return an agenticbackofficeaudit with readiness, replayresults, adversetests, controlfindings, systemintegrity, oversighthealth, recoverytest, metrics, decision, requiredrepairs, owner, and nextreview.

Build a Decision Evidence Packet

Use this workflow when preparing a review-ready evidence packet for a finance, procurement, legal, people, compliance, or administrative decision while preserving source provenance, uncertainty, policy boundaries, and specialist judgment.

Output Return a decisionevidencepacket with object, decision, requester, owner, reviewer, facts, evidenceinventory, policyorplaybookmatches, missinginformation, conflicts, riskflags, recommendation, proposedaction, approvalrequired, sourcecitations, and auditfields.

Design a Back-Office Loop

Use this workflow when converting a recurring finance, procurement, legal, people, compliance, or admin request from a ticket queue into a governed operational loop with explicit objects, states, evidence, owners, and actions.

Output Return a one-page backofficeloopspec with object, candidatescore, states, transitions, evidence, owners, sourcesystems, allowedreads, allowedwrites, approvalgates, exceptions, auditevents, rollback, killswitch, metrics, and loopowner.

Map Operational Truth

Use this workflow when an internal workflow reads across fragmented systems or risks creating a shadow source of truth, and operators need explicit field ownership, conflict rules, identity attribution, and authoritative write paths.

Output Return an operationaltruthmap with rows for object, field, authoritativesystem, owner, sourceclass, allowedreads, authoritativewrite, freshness, conflictrule, reconciliationowner, identityevents, and prerequisitegaps.

Set Delegation Depth

Use this workflow when deciding how far an agent may prepare, recommend, draft, execute, or self-review a back-office workflow based on reversibility, materiality, evidence confidence, reviewer capacity, and observed reliability.

Output Return a delegationmatrix with workflowclass, action, riskdimensions, mode, threshold, requiredevidence, reviewer, reviewreason, log, rollback, promotioncriteria, downgradetriggers, and reviewdate.

Human review

Keep consequential decisions explicit.

These workflows structure analysis and artifacts; they do not authorize autonomous external, personnel, financial, legal, security, or other high-impact action.

Source and files

Grounded in the complete Agentic Back Office series.

10 source essays · 5 usable skills · reviewed Jul 2026

Choose your reading rhythm.

Start with a weekly briefing, add daily notes, or hear only when a durable essay or library update is ready.

You've successfully subscribed to Antoine Buteau
Great! Next, complete checkout to get full access to all premium content.
Welcome back! You've successfully signed in.
Unable to sign you in. Please try again.
Success! Your account is fully activated, you now have access to all content.
Error! Stripe checkout failed.
Success! Your billing info is updated.
Error! Billing info update failed.