Work & experiments
ExperimentInternal workflow trial

Brand Launch Agent

Can one approved brief carry an idea through naming, validation and a coherent design handoff?

Latest update · Used the workflow in a real naming exercise
My focus
Workflow design and internal evaluation
Format
Agent skill
  1. 01Approve the brand brief
  2. 02Explore name candidates
  3. 03Validate with evidence
  4. 04Prepare the design handoff
Workflow exercised internally; independent-user validation remains pending.

Why this exists

Naming a media project meant moving between ideas, domain checks, social handles and design. A name could sound right and still be unusable. Each handoff meant explaining the same idea again.

That experience suggested a product question: could an approved brief preserve the context through the whole process while keeping uncertain availability checks honest?

What I’m testing

The first version is an agent-skill workflow, supported by schemas, templates and narrow helper scripts. An internal retailer-protection naming exercise carried an approved brief through candidate comparison and a visual-design handoff.

The broader direction includes a design handoff and launch package. Account claiming remains user-guided.

Key decisions

  • Start with a skill. Test the workflow before investing in a standalone application, accounts and billing.
  • Treat availability as evidence. Retain the exact name checked, method, timestamp and uncertainty with every observation.
  • Keep approval boundaries explicit. The user approves the brief and selects the name. An unclear result must not silently become an available name.

How it works

The skill coordinates the sequence. Deterministic scripts handle tasks such as normalising names and validating evidence records. Connected tools would supply live checks where a reliable method is established.

The intended output includes the approved brief, candidate rationales, validation observations, selection and design handoff. A web application is a later option if actual usage demonstrates a need for persistent history or collaboration.

Evidence and learning

The CloneBrake exercise shows how the checks affected a decision. Cloneward was rejected after exact social-name conflicts; Impostra had a registered .com and occupied handles, including an adjacent fraud-content channel. CloneBrake survived the initial checks and I confirmed purchasing its .com domain.

The checks also preserved uncertainty: no visible account did not mean a handle could definitely be registered, Facebook could not be checked reliably, and trademark research was incomplete. This is evidence of an exercised comparison and selection workflow, not blanket name clearance or validation with independent users.

The shift from an initial web-app proposal to a skill-first approach is itself a useful scope decision: the first question is whether the workflow helps someone make a better decision.

Project evolution

Recent decisions and releases, newest first.

  1. implementation

    Used the workflow in a real naming exercise

    Carried CloneBrake through an approved brief, name comparison, availability checks and a visual-design handoff.

  2. Architecture decision

    Chose a skill-first starting point

    Reduced the first delivery scope to a reusable skill before building a web app.

  3. Product definition

    Defined the brand launch workflow

    Defined a connected journey from brand brief to launch package.

View full project history (3) →

What’s next

Test the validation methods and finish the unresolved checks. Then observe whether another person can complete the workflow without my guidance, recording failures and uncertainty rather than treating an internal trial as a launch.

Play by Value provided the initial naming context. CloneBrake became a concrete internal trial, connecting the brief, candidate checks, selected name and design handoff.