Agent Setup & Workflows

Claude Code & Codex.
Configured for your work.

An established development setup adapted to your projects, with specialist agents, reusable skills, organised context and workflows for research, design and implementation.

Your task

You want a working environment that starts with the relevant project context and repeatable procedures, so less of each task is spent reconstructing instructions and setting up tools.

What you receive

A configured environment, project context pack, specialist roles, skill library, tested workflows and editable configuration with operating notes.

Worked delivery example · This website

Context, specialist tasks,
review and a working result.

The Adel Labs website update follows the workflow below. Explore the execution record; the controls here present the record and do not start a model or access your files.

Context

Start from the current project.

The website requirements, project inventory, approved LINK mark and existing recordings established what had to be retained. The audit supplied specific changes and acceptance checks.

Explore the implemented project collection →
Permissions, connections and recovery
  • The scope names accounts, folders, permitted actions and external data routes.
  • Drafting and sending are separate actions. Messages and changes follow the agreed review or prior permission.
  • Failed inputs and providers produce a visible status. The agreed retry or recovery procedure must not duplicate an external action.
Scope, deliverables and acceptance checks

Scope and deliverables.

Your proposal identifies the included deliverables and their checks. The outline below helps define that scope.

  • Environment setup: configure the agreed Claude Code or Codex workspace, supported tools and permitted connections around your existing projects.

    Acceptance criteria

    Complete a representative task in the configured project using the agreed tools and access.

  • Project context: organise requirements, file maps, working conventions and current decisions so each task starts from the relevant material.

    Acceptance criteria

    Start a fresh task and confirm that current requirements, source files and project decisions are identified.

  • Specialist agents: define research, design, implementation and review roles, with task routing and handover suited to the selected platform.

    Acceptance criteria

    Follow a task through the agreed roles; inspect each output and the review at handover.

  • Reusable skills: adapt established procedures for recurring outputs, including inputs, templates, checks and file destinations.

    Acceptance criteria

    Repeat a workflow with a second input and compare the result with its stated checks.

  • Workflow optimisation: run a representative task set, refine context and task sequences, and record observed differences in completion, review effort and resource use where measurable.

    Acceptance criteria

    Compare the current process with the configured workflow on representative tasks.

  • Connected operations: configure approved document intake, reporting routines and integrations with named accounts, outbound data and review steps.

    Acceptance criteria

    Inspect the permitted actions and external data routes; demonstrate review before an external change.

  • Handover: provide editable setup and workflow files, a worked example, operating notes and a documented update or recovery procedure.

    Acceptance criteria

    Restore the delivered configuration, run the worked example and confirm that another authorised user can follow the notes.

First engagement

A scoped brief and, where needed, one working prototype. Review the task, content and result before extending the build.

At handover

The proposal names the editable source, configuration, assets and operating notes you receive. Support and maintenance are agreed separately.

What sets the fee and schedule?
  • The selected platform, project scope, specialist roles and existing configuration.
  • The skill library, workflow complexity, integrations and representative task set.
  • Review steps, provider costs and the recovery behaviour required.

These inputs determine the proposal. No fee or delivery date is fixed before the included work and dependencies are agreed.

Scope, exclusions and handover

Subscriptions, provider usage, new integrations and maintenance are identified separately. Provider-backed configurations use connected model services. Fully disconnected work requires a separately scoped local model system and acceptance test.

The proposal identifies the source, configuration, access and operating notes included at handover. Third-party software and model licences remain applicable. Maintenance, update testing and further development are scoped separately.

Before we begin

Questions about this scope.

Can it work through Telegram or email?

Yes, as a connected workflow. Messages and email pass through their providers. The scope states which tasks, excerpts or status messages leave the local system; these connectors are not part of the offline workflow.

Read the permissions and review scope →
What happens after launch?

A one-time build includes agreed acceptance checks and a walkthrough. Maintenance, update testing, support allowance and further development can be scoped separately.

How does a project begin?

Prepare a brief with the task, the tools you use and the required result. We agree the included work, fee, schedule and review points before implementation. The first engagement can be a scoped prototype when the full build still needs to be defined.

Agent Setup & Workflows

Configure your development workflow.

Bring an existing project and a task you repeat. Define the environment, reusable skills, specialist roles and review criteria around that work.

Prepare a project brief ↗

Agents, skills & working methods

Specialist AI agents.
Built around your workflow.

Equip your team with AI agents for marketing, coding, design and research, configured around your brand, documents and working methods. Reusable skills provide task-specific procedures, templates and checks. For larger assignments, coordinated agents divide the work, review results and prepare a connected handover.

Marketing and search

Brand positioning and voice; search audits and content briefs.

Coding and product development

Debugging, interface implementation and site checks.

Design and creative production

Art direction, interface design, campaign and product content.

Research and verification

Source analysis, reference verification and numeric audits.

Reports and presentations

Technical writing, workbooks, presentations and interactive reports.

How agentic workflows work

Agentic AI carries out a sequence of tasks with tools and checks along the way. A skill supplies a reusable method; an agent applies it; a coordinated team divides a larger assignment between specialist roles.

Each setup uses a selected skill set, tested on representative tasks. Setup, testing and handover are included in the agreed scope; continuing maintenance is scoped separately. Connected development tools and fully offline model workspaces are separate configurations.