NewEvery agent tool call now routes through the governed MCP gateway — policy check, audit log, then dispatch.Read the changelogarrow_forward
XOOZLYAI Orchestrator
routeHow it workssmart_toyAgent fleetshieldGovernance & trustsellPricingmenu_bookDocsarticleBloghistoryChangelog
Log inStart free
grid_viewProductsrouteHow it workssmart_toyAgent fleetshieldGovernance & trustsellPricingmenu_bookDocsarticleBloghistoryChangelogHelp centreLog in
Start free
  1. XOOZLY
  2. chevron_rightDocumentation
Loading the knowledge base…
XOOZLYAI Orchestrator

Ready to orchestrate your AI fleet with total confidence?

Start a project, set a cost cap, and watch the loop run with every tool call policy-checked and logged before it leaves the platform.

Start freearrow_forwardSee what shipped
  • check_circleNo credit card required
  • check_circleYour keys, your model spend
  • check_circleEvery tool call audited

Platform

  • All productsarrow_forward
  • Orchestration looparrow_forward
  • Agent fleetarrow_forward
  • Governed control planearrow_forward
  • Pricingarrow_forward

Trust & legal

  • Trust centrearrow_forward
  • Security & compliancearrow_forward
  • AI use policyarrow_forward
  • Report a vulnerabilityarrow_forward
  • Data subject requestsarrow_forward

Resources

  • Documentation & guidesarrow_forward
  • Engineering & research blogarrow_forward
  • Changelogarrow_forward
  • Uptime & reliabilityarrow_forward

Browse by domain

  • Strategy & Researcharrow_forward
  • Developmentarrow_forward
  • QAarrow_forward
  • Governance & Orchestrationarrow_forward
  • All 10 domainsarrow_forward
XOOZLY
© 2026 XOOZLY · 61 agents · 10 domains · 8 SDLC stationsAI use policyarrow_forwardLog inarrow_forward99.9% uptime SLAarrow_forward
  1. XOOZLY
  2. chevron_rightDocumentation
  3. chevron_rightGuide
Loading this guide…
  1. XOOZLY
  2. chevron_rightDocumentation
  3. chevron_rightHow rule files govern an agent, and how a version is approved
arrow_backBack to the knowledge base
schedule1 min read·verifiedChecked against the code on 25 October 2026
Core concepts & getting startedrule-files-and-versionscodeapi-server/domains/rules_optimization/service.py

Agent configuration

How rule files govern an agent, and how a version is approved

An agent's behaviour lives in its rule file, and a rule file is never edited in place. Every change writes a new version row, which is what makes a rollback a selection rather than a reconstruction.

open_in_newOpen Agent fleet in the consolearrow_forward
01

Versions accumulate; they never overwrite

An agent's rule file is the text that governs how it behaves. It is never edited in place: a change writes a new version row and the previous one stays exactly as it was. That is the whole mechanism behind a rollback being a selection rather than a reconstruction.

historyWhy append-only matters hereBecause versions accumulate rather than overwrite, the question “what was this agent running on the day of that incident?” has an answer. An in-place edit would have made it unanswerable.
02

What creates a new version

Two things write a new version: a person editing the rule file directly, and a person approving a Rule Optimizer proposal. Both end in the same place — a new row — and neither happens without someone deciding it should.

03

What a rule file holds

A rule file is the text that governs how one agent behaves — what it should do, what it must not do, and how it should format what it produces. It is content, not configuration: there is no schema to satisfy, which is why reviewing a change means reading it.

descriptionSome agents carry more than oneA single agent can have several rule files covering different parts of its behaviour. Don't assume one agent means one file when you are looking for the text behind a decision.
04

Rolling back

Because a change writes a new row rather than editing one, rolling back is selecting an earlier version — not reconstructing it from a diff or a backup. Every version that ever applied is still there, with the time it applied from.

  • The question “what was this agent running when that run executed?” has a direct answer.
  • A rollback is itself a new version, so the history stays append-only rather than rewinding.
  • Nothing is deleted on rollback, so rolling forward again is the same operation in the other direction.

boltPut it into practice

Run a fleet with these guardrails on from day one.

Hard cost caps, one audited gateway in front of every tool call, and versioned rule files — the defaults, not an enterprise add-on.

Start freearrow_forward

On this page

4 sections
  1. 1. Versions accumulate; they never overwritechevron_right
  2. 2. What creates a new versionchevron_right
  3. 3. What a rule file holdschevron_right
  4. 4. Rolling backchevron_right
feedbackSomething missing?

Related guides

  • articleSetting a per-project cost capSet a cost cap
  • articleConnecting your first MCP serverConnect an MCP server
  • articleUnderstanding run states and lifecycle transitionsRun log