WeXL Forge · How we build

Accelerators for every stage of the build.

Forge is our collection of AI agents, one for each step from spec to production, and the Workbench we set up inside your teams to run them. The agents take the repetitive work; our senior engineers make the calls and check every output. Each agent below is published with a worked example and proof you can check.

The Forge Catalog

11 agents

07
Run
Built on demand
08
Govern
Built on demand
The WorkbenchRuns on Kiro, Claude Code or OpenCode, with any frontier or open-weight model.
LiveBetaBlueprint
Stage
Status
Proof

The Workbench

Installed in your first two weeks. Yours to keep.

01

Steering & standards

Your team's conventions, architecture rules and definitions of done, encoded so every agent follows them.

02

Skills library

Reusable, tested instructions for the jobs your squads repeat: specs, APIs, migrations, reviews.

03

Role-based agents

BA, developer, reviewer and tester agents that work inside your repositories and pipelines.

04

Guardrails & review

Hooks, CI checks, automated code review, and policy checks with human approval on sensitive agent actions, all recorded, so quality holds at agent speed.

05

Evals & observability

Quality, drift and cost of every agent in production, tracked and reported.

06

Playbooks & training

Your people learn the method and keep running it after we step back.

Runs onAWS KiroClaude CodeOpenCodeFrontier modelsOpen-weight models

Three ways to use an agent.

In your engagement

Forge comes with every WeXL delivery. Nothing extra to buy.

Under your brand

Services firms can ship any agent white-labelled, inside their own platform.

Built to own

Have an agent built for your workflow and take the source.

Start here

See Forge on your own codebase.

Start a conversationWe'll show the Live agents on a real workflow of yours.