What happens after
your agent finishes?

You read the output. Then the session ends, and next week the same agent finds the same things and argues the settled ones again.

$ npx @uluops/setup

TRACKER→REGISTRY→

Scroll

What if the agent could see that you already decided that one was a false positive, and why?
What if the same defect came back after a fix, and you knew it was the same one?
Which of your agents actually find things that matter?
What if decisions are being made faster than anyone writes them down?

Most of those are things UluOps does today.

The operations layer for agentic work.

Git remembers what changed. UluOps remembers what was decided — what was decided about what the agents found: fixed, deferred, wontfix, false positive, or noted and left alone, each with its reason.

255 agents

6 agent types

4 definition languages

The System

Four layers. One feedback loop.

Each layer feeds the next. Definitions become runs. Runs become data. Data improves definitions. The system gets smarter with every run.

Registry

DEFINE

Define AI agents in structured YAML. Version them. Publish to a shared registry. 255 agents across validators, analysts, explorers, and forecasters.

Tracker

VALIDATE

Run agents against any artifact. Track issues with fingerprint correlation. Detect regression automatically. See quality converge over time.

Improvement

IMPROVE

Agents audit agents: meta-layer agents review domain agents. Improvements cascade through the ecosystem — the property the research page calls recursive appreciation.

SDK & MCP

INTEGRATE

Integrate into any workflow. MCP servers give Claude Code native access. The SDK handles orchestration. One command: npx @uluops/setup.

Two Interfaces

Define. Run. Decide.

app.uluops.ai

++++

▶OPS-ULUOPS-REGISTRYDASHBOARD

3open

14fixed

2deferred

[ VALIDATORS ]

↑code-validator92

→security-analyst88

↓test-architect78

registry.uluops.ai

code-validatorv2.1.0

PUBLISHEDCOMMUNITYsoftware

Validates code quality, standards compliance, test coverage, and best practices. Blocks progression if critical issues found.

Scoring0–100 · 6 dimensions · 48 checks

Executions1,247 runs · 12 projects

Authoruluops

Execution History1,247

code-qualitytypescriptsecurity

ONE RECORD

The Loop

Self-improving by design

Every validation run closes the feedback loop. The system doesn't just measure quality — it compounds it.

{ }

DefineRegistry

ValidateAgents

TrackTracker

ImproveRAH

{ }

DefineRegistry

ValidateAgents

TrackTracker

ImproveRAH

↻ CONTINUOUS LOOP

Multi-Target

One command. Multiple harnesses.

Install UluOps into every agentic harness on your machine in a single run. Failure isolation, per-target summary, one set of prompts.

$ npx @uluops/setup --all-detected

Claude Codestable

OpenCodestable

Gemini CLIstable

Codexexperimental

Or pick a subset: --harness claude-code,gemini-cli. Auto-detection skips experimental harnesses; opt in to Codex with --harness codex.

How It Works

Three steps. Thirty seconds.

Install, validate, track. Every run feeds the system. Quality compounds from the first command.

install

$ npx @uluops/setup --signup

  ⟨u⟩ ulu·ops
      the operations layer for agentic work

  ✓ Account created (you@example.com)
  ✓ API key generated
  ✓ MCP config → ~/.claude.json
  ✓ 23 agents · 28 commands installed

  Setup complete!

1,687

Runs examined

29,253

Findings tracked

19,055

Marked fixed

579

Published definitions

What came back, January 2026 to 2026-10-22

8: Caught by re-running: a later run found a resolved finding again
6: Reopened by decision: a resolved finding reopened by a person or an agent, with no run involved

Counts, not rates, and nothing here says how many were missed.

Our own record, not a benchmark: one operator, 40 weeks, 29,253 findings tracked. Read from the tracker, not typed.

as of 2026-10-22 · 255 of the 579 published definitions are agents

For Developers

SDK-first design

Five packages. One pipeline. The SDKs handle auth, type safety, and error handling. The CLI ties them together.

Submit validation runs, track issues, query analytics. The operational layer.

$ npm install @uluops/ops-sdk

example.ts

import { OpsClient } from '@uluops/ops-sdk';
 
const client = new OpsClient({
apiKey: process.env.ULUOPS_API_KEY,
});
 
// Submit a validation run
const run = await client.runs.submit({
project: 'my-app',
agent: 'code-validator',
score: 87,
issues: findings,
});
 
// Query project analytics
const trends = await client.analytics.getTrends(
'my-app'
);

Pricing

Start free. Pay for room.

The catalog, the issue lifecycle and every analytics view are free on every tier. Tiers meter workspace: projects, private definitions, members. Agents generate more than anyone can hold in their head, and the audits and the decisions about them need somewhere to sit and scale.

On every tier, including Free

All 579 published definitions — 255 agents, 226 commands, 75 pipelines, 23 workflows
Findings tracked as issues, fingerprinted across runs, regressions caught
Every decision on the record with its reason — fixed, deferred, wontfix, false positive, observation
Every analytics view: agent reliability, trends, taxonomy, cost, cross-project
MCP servers, CLI and SDKs — bring your own model key