Live Sundays at 2 a.m.

Custom scripts audited weekly — findings drafted, nothing touched.

Code Inspector

Once a week, in the small hours of Sunday, it reads every custom script and field in your system and drafts a findings report — security risks and costly query patterns, each finding quoting the offending line. It changes nothing; your developer decides what to fix.

Like every Infinary agent, it drafts — a person on your team approves before anything is posted.

What it does

The Code Inspector audits the customizations layered onto your system: server scripts, client scripts, and custom fields. It looks for injection risks, unsafe evaluation, missing permission checks, secrets left in code, and the query patterns that quietly get expensive — unfiltered full-table reads, loops that query per row. Every finding quotes the line it's about and names the script it lives in, with severity judged by consequence. The report is a draft; the agent is read-only on your code and never edits, disables, or deploys anything. A script it cannot read is listed as unaudited, by name.

The facts on Code Inspector
Status
Live
Runs
Sundays at 2 a.m.
Cost per run
Up to $3
Review
Creates drafts only — you approve before anything is posted.
What it reads
your existing reports, your records, live totals and history

The Problem Today

"Customizations accumulate. A script written in March quietly scans a whole table by June, and nobody reviews code that already works — until the day it doesn't."

How Infinary Handles It

A standing weekly review, done the same way every time. Findings arrive as a drafted report with the exact lines quoted — your developer stays the one who changes code.

What a run looks like

What a run looks like

Say a custom inventory-sync script works fine in testing but hides a loop that fires a query per row. On Sunday's run, the report quotes that exact loop, names the script, and rates the finding by what it would do under a real order surge — lock the tables everyone else is using.

Monday morning, the developer reads the drafted report, reproduces the problem in staging, and rewrites the loop as one batched query. The agent never touched the code; it made sure a person did.

INF-SCENARIO-TRACE // code-guardian Core Platform
Custom scripts audited weekly — findings drafted, nothing touched.

Code Inspector

DEPLOYED IN CLIENT PRIVATE CLOUD
DATA_RESIDENCY: SOVEREIGN REVIEW: HUMAN_APPROVED
Example Scenario CODE-GUARDIAN-SCENARIO-01

How It Works

Step
By Step

01

Read

It reads every server script, client script, and custom field in your system.

02

Analyze

It checks for security flaws and costly query patterns, judging severity by consequence.

03

Draft

It drafts a findings report, each finding quoting the offending line and naming its script.

04

You approve

Your developer reads the report and decides what gets fixed, and when.

What Changes

What You Get

Coverage Every script, weekly

The whole customization surface is re-read every week — including the code nobody has opened since it shipped.

Trail Line-level

Every finding quotes the offending line and names the script it came from.

Boundary Read-only

It never edits, disables, or deploys code — fixes are made by your developer.