NewKnowledge base governance playbook is live
← Back to work
Case studyLocal-first support tool

ChartFinder

Chrome extension for cleaner support context capture

A local Chrome extension that copies visible shipment, client, and support article context into a clean format for support replies.

The tool addressed a queue problem: teammates needed clean context quickly and a data model they could explain and review.

Chrome extensionLocal firstJavaScriptClipboard workflowSupport formatting
In brief

Replace repeated copying with one reviewable action

ChartFinder reads the context already on screen, formats it for a support reply, and keeps the process local and reviewable.

ChartFinder extension

Current page

Support context

Output

Markdown or HTML for support replies.

Data boundary

Visible page text, clipboard, local preferences.

Review reminder

Double-check copied details before sending.

Public-safe mockup showing the extraction and clipboard pattern.

Screenshot

See the tool in context

A public-safe view of ChartFinder copying article context into a clean reference.

Support article capture screenshot
Support article capture

ChartFinder copies a support article into a reusable reference for a support reply.

Technical shape

Match the implementation to the data boundary

The extension reads visible browser content, copies output to the clipboard, and stores preferences locally.

Tool type

Chrome extension

Stack

JavaScript + HTML/CSS

Output

Markdown / HTML

Storage

Chrome local storage

Problem

Support work included too many small copy-and-paste steps

Support teammates repeatedly copied tracking numbers, shipment pages, service details, client links, and article references. Each step was small, but the repeated work slowed the queue.

Constraint

Set a data boundary people could understand

The extension handled support context, so its data model had to be easy to explain. It read visible page text, ran in Chrome, made no backend queries, sent no data externally, and stored preferences locally.

Approach

Keep the tool inside the existing browser workflow

A support rep could open a shipment, client page, or article, then copy the relevant context as Markdown or HTML. The output stayed visible and editable before use.

Outcome

Document the tool so others can use and maintain it

ChartFinder grew from a copy helper into a documented workflow with setup steps, troubleshooting, known issues, compliance notes, feedback, and a changelog.

Supported contexts

Support three common copy tasks in one tool

The extension handles shipment details, client context, and article links on pages teammates already use.

Context type

Shipment details

Extracted visible shipment context like recipient, service, insurance, tracking number, cost, and related shipment page links.

Context type

Client details

Copied client context into a cleaner format, including a linked client name and available email context for support handoffs.

Context type

Support articles

Copied support article or internal resource context into a reusable reference format, including an internal-only reminder when needed.

Workflow

Open, extract, format, and review

The workflow stays in the browser and avoids another destination.

Stage 0101

Open the work context

Start from an authenticated support page, client page, support article, or internal resource already visible in the browser.

Stage 0202

Extract visible text

Read the information available on the current page, without backend account access or external data calls.

Stage 0303

Format for support

Format the selected context as Markdown or HTML, then copy it to the clipboard.

Stage 0404

Review before sending

The support teammate reviews the copied output before using it in a customer or internal message.

Data boundary

Keep the data flow local and visible

The guide separates visible text, clipboard output, local preferences, and external transmission.

Boundary

Visible text only

ChartFinder worked from the current page surface. It did not search customer accounts or reach into backend systems.

Boundary

Local operation

The extension ran inside Chrome, with copied output going to the local clipboard and preferences stored in Chrome local storage.

Boundary

No external transmission

The documented data flow avoided API calls and external communication, which made the tool easier to explain and review.

Boundary

Human review stays in the loop

The guide told users to double-check copied output, especially around known parsing edge cases like service or insured-value detection.

Adoption

Support the tool with clear documentation

The guide covers setup, use cases, troubleshooting, known issues, compliance notes, feedback, and changes.

Adoption layer

Setup instructions

Walked teammates through Chrome extension setup, developer mode, extension loading, pinning, and update flow.

Adoption layer

Usage guide

Explained shipment lookup, client information copy, article copy, auto-detect behavior, and output format choices.

Adoption layer

Troubleshooting

Explained common messages about refreshes, missing details, empty copies, and HTML formatting, with a next step for each.

Adoption layer

Known issues

Documented cases where service or insured-value detection still needs review.

Iteration

Use the changelog to show what improved

Updates improved reliability, support formatting, and preference controls.

v1.5.3

More reliable injection and clear success feedback

Improved long-lived tab support, added clear success feedback, and shortened ping timeouts without changing the local data model.

v1.5.2

Autorun and support formatting

Added autorun, shipment detection, and Markdown output for faster support replies.

v1.5.1

Preferences panel and context controls

Added preferences for output and behavior.

v2.0.0

Broader toolkit direction

Outlined a broader support toolkit while keeping the current tool focused.

Result

Automate a support task without hiding the boundary

The work combines frontline support, browser automation, privacy-aware design, and adoption documentation.

Lesson

Small repeated tasks deserve design attention

Copy-and-paste work can look too minor to fix. In a busy queue, repetition makes it worth addressing.

Lesson

A clear local data model supports trust

The model is simple to explain: visible text in, formatted clipboard output out, preferences stored locally.

Lesson

Documentation supports adoption

Setup, troubleshooting, compliance notes, feedback, and a changelog made the tool usable beyond its first release.

ChartFinder reduced a repeated queue task with a local, reviewable, documented browser tool.