The Naming of Fabriq: From dapei to Workshop

TL;DR. Fabriq's name evolved through three iterations, each reflecting a deeper understanding of the problem. From "dapei" (搭配 — composing parts into a whole) through "fabric" (the woven weave) to "fabriq" (the workshop where weaving happens). This post traces that evolution and the philosophy each name captured.

---

大培 — dapei#

The project started in mid-2024 under a Chinese name: 大培 (dà péi), romanized as dapei.

The characters were chosen deliberately:

• 大 (dà) — large, great, comprehensive
• 培 (péi) — to cultivate, to nurture, to train

Together they evoke the idea of a comprehensive cultivation environment — a system that doesn't just edit code but grows the capabilities of the team using it. The name reflected our early conviction that AI tools should teach as much as they do.

In practice, dapei shipped as a headless engine and a VS Code extension. It proved the technical model — CDR, evidence-first artifacts, the scanner-validator pipeline — but the name always carried a friction. Outside Chinese-speaking contexts, dapei was hard to pronounce, hard to search, and carried none of the semantic weight it had in the original language.

When we started building the desktop shell, we knew the brand needed to travel further.

Fabric#

The first English iteration was fabric — a direct translation of the conceptual thread.

The metaphor worked well:

• A fabric is woven from individual threads — just as software is woven from individual decisions, commits, and components.
• A fabric has visible seams — the weave is legible if you look closely, and Fabriq's job is to make the weave legible by default.
• A fabric is flexible yet strong — it adapts to the shape of what it wraps.

The name captured the output: a system that turns AI actions into an auditable, legible whole.

But it missed something important. It described the result, not the process. A fabric is what you get after the work is done. The name didn't communicate the active, messy, iterative process of building software.

Fabriq#

Fabriq is a deliberate misspelling of fabric — but more than that, it's a shift in meaning.

A fabriq (in the archaic sense) is a workshop — the room where raw materials are cut, joined, shaped, and assembled into something that holds together. The forge. The workbench. The place where making happens.

The shift from fabric to fabriq is:

Aspect	Fabric	Fabriq
Focus	The output	The process
State	Finished	In progress
Role	Result	Workshop
Implication	"Here is what was made"	"Here is where making happens"

Fabriq is the place where the human's intent and the agent's actions get woven into a working product — with the seams visible throughout the process, not hidden in the final artifact.

The Chinese layer#

"Fabriq" also works in a bilingual context. The character 培 (péi) from the original dapei name shares a conceptual root with "workshop" — a place where things are cultivated through deliberate effort.

The project's documentation and code comments alternate between English and Chinese naturally, reflecting the team's operating language. The name "Fabriq" was chosen to read naturally in both:

• In English, it reads as a distinctive brand — recognizable, searchable, memorable
• In Chinese, it suggests 工坊 (gōngfáng, workshop) — the same workshop metaphor, familiar and concrete

This duality matters because the project inherits from an engineering tradition that crosses both languages. The dapei-skill repository contains documentation in both. The Fabriq codebase continues that practice. The name bridges them.

What the name commits us to#

Naming a product after a workshop sets an expectation: this is a place of making, not a place of magic.

A workshop commits to:

Expectation	What it means
Visible process	You can see how the work is progressing at any moment, not just when it's done
Clear boundaries	Tools have designated places; work-in-progress is separated from finished work
Tool transparency	Every tool on the wall has a known purpose and known effect
Cleanup as workflow	Closing a feature is as deliberate as opening one — the workshop returns to a clean state

These aren't UX metaphors — they're architectural principles that shaped Fabriq's design. The two-dimensional workspace/feature split, the confirmation-gated IPC, the evidence-first artifact pipeline — all of them can be traced back to the question "what would a well-run workshop do?"

The lineage continues#

The original dapei-skill repository is still the source for several core algorithms — Tree-sitter integration, CodeGraph, evidence validation, CDR pipeline stages. Fabriq replays those algorithms through a desktop shell: the same scanner-validator pipeline, same three-level evidence scale, same symbol-anchored provenance.

The name changed, the architecture evolved, but the engineering lineage is continuous:

dapei (proved the model) → fabric (named the output) → fabriq (built the workshop)

Each name captured a real insight. The current one captures the deepest: that the most important thing an AI-assisted development tool can do is not generate code — it's provide a workspace where generation and verification happen in explicit dialogue.

That's what a workshop is. A place where making and inspecting are the same process.

---

The workshop is open. Download Fabriq or read the developer guide to start building.