MergeLoom

Fix The Work Nobody Has Time For

Give security hygiene, docs drift, dependency cleanup, and platform standards their own scoped agents. They open review-ready PRs/MRs, not chaos.

Beta. More codebase care, less sprint tax.

BetaCodebase AgentsCustom MandatesRepository ScopeDaily BudgetsFile PoliciesQuality ControlsHuman Review
Fleet Loop
Product Map
  1. Step 01
    Set Mandate
  2. Step 02
    Select Scope
  3. Step 03
    Budget Runs
  4. Step 04
    Open Review

Stop Maintenance Stealing The Sprint

Small codebase issues are cheap early and expensive later. They just rarely beat roadmap work.

Small Fixes Pile Up

Security, docs, dependencies, and standards drift until they hurt.

Product Time Gets Taxed

Engineers lose roadmap time to repeatable cleanup.

One-Off AI Fades

The mandate, budget, scope, and evidence vanish after the session.

Review Noise Scares Teams

Nobody wants automation that floods reviewers.

Narrow Agents. Useful PRs/MRs.

Agent Fleets are named Codebase Agents with a mandate, scope, budget, file rules, cadence, and review boundary.

Active Agents
4
Daily PR Cap
6
Open Review Cap
8
Merge Control
Human
Active Agent Mandates

Specialists With A Narrow Job

Each agent owns one recurring concern. No broad autonomous wandering.

Security Agent
payments/*
Daily

Finds safe config and vulnerable-pattern fixes.

Docs Agent
api + setup
3x week

Repairs docs drift around changed services.

Dependency Agent
service packages
Weekly

Moves small compatibility cleanup into review.

Standards Agent
platform rules
Daily

Applies repeatable engineering standards.

Control Plane

Mandate, Scope, Budget

The fleet can be proactive without flooding the team. You set the work boundary before it runs.

Human Review
Repo And File Scope

Choose the repositories, include paths, exclude paths, and ownership rules.

Bounded
Daily Output Limit

Cap PR/MR-producing runs so reviewers never get swamped.

6/day
AI Usage Budget

Keep proactive work inside credit, token, or self-hosted provider limits.

Capped
Quality Path

Context Engine, validation, repair, AI review, and audit evidence still apply.

Required
Review Queue

Only Useful PRs/MRs

Agents open focused fixes when the mandate is met. Humans still review and merge.

Security config cleanup

Small fix with validation and risk notes attached.

Security AgentChecks passed
API docs drift

Docs update linked to the changed service contract.

Docs AgentEvidence saved
Package compatibility

Dependency cleanup capped to a reviewable branch.

Dependency AgentHuman review
Fleet Mode
Scoped agents
Spend Control
Daily budget
Review Load
Open cap
Audit Trail
Every run

More Codebase Care. Less Sprint Tax.

Keep the codebase improving without pulling engineers away from product delivery.

Fix Security Lag

High-confidence fixes keep moving.

Encode Standards Once

Platform rules become repeatable reviewable work.

Cap The Cost

Daily budgets keep proactive work predictable.

Keep Evidence

Mandate, scope, budget, checks, and PR/MR stay linked.

Old Way

  • Security Waits
  • Docs Drift
  • One-Off AI
  • No Review Cap

Agent Fleets

  • Named Mandates
  • Scoped Repos
  • Daily Budgets
  • Quality Gates

Scope, Budget, Evidence

Agent-created jobs use the same MergeLoom path: scope, checks, usage, audit, and PR/MR review.

MergeLoom Repository Catalog showing repository routing, aliases, labels, and controls used to scope automated codebase work.
Repository scope controls where a Codebase Agent works.
MergeLoom Quality Agents timeline showing clarity, investigation, validation, repair, review, Diff Guard, and PR or MR handoff evidence.
Fleet-created jobs still get checks before review.

Control The Fleet

Useful automation needs limits. Agent Fleets are bounded before they create review work.

  • Repo And File Scope

    Selected repositories only. Include and exclude rules available.

  • Daily PR/MR Budget

    Limit PR/MR-producing runs per day.

  • AI Usage Budget

    Cap Cloud credits or Self Hosted tokens.

  • Open Review Control

    Limit review load. Pause or resume any agent.

Keep Recurring Work Moving

Security, docs, dependencies, tests, and standards keep moving in small reviewable PRs/MRs.

Security Hygiene

Keep high-confidence fixes moving.

Documentation Drift

Keep docs closer to shipped code.

Dependency Cleanup

Move small config and compatibility fixes.

Test Gaps

Add focused tests around stable behavior.

Platform Standards

Clean up recurring patterns across services.

Maintenance ROI

Turn cleanup into measured output.

Agent Fleets FAQ

What Are Agent Fleets?

Scoped Codebase Agents for recurring codebase work. Each has a mandate, repo scope, file rules, budgets, and PR/MR review path.

Will They Flood Reviewers?

No. Cap PR/MR-producing runs, AI usage, open reviews, repository scope, and pause/resume controls.

Do They Need Jira Tickets?

No. They create normal MergeLoom jobs when the mandate and scope justify reviewable work.

What Keeps Them On Mandate?

Scope, file rules, budgets, Quality Agents, validation, audit evidence, and human review sit above the agent mandate.

Explore Related Product Areas

Start Free With No Risk

Pay For Outcomes, Not Seats

Run MergeLoom on scoped work before rolling it out. You only pay when a run opens a PR/MR for review, not for seats or tickets that stop before handoff.

Cloud

50 Free PR/MR Runs

Then From £4 Per PR/MR

Self Hosted

50 Free PR/MR Runs

Then From £2 Per PR/MR

Paid Outcomes

Only PR/MR Runs Count

No PR/MR, No Run Charge

Start Free Request A Free Demo
  • Free To Start
  • Pay For Outcomes
  • No Lock-In Contracts
  • No Credit Card Required (Self-Hosted)
  • Cancel Anytime

No PR/MR, No Run Charge · No Seat Pricing · Human Review Stays In Control

See Pricing