← All runbooks
gooseworks-ai / capabilities-brand-voice-extractor

Brand Voice Extractor

Analy

agent codexmodel gpt-5.5snapshot python312-uveval programmatic7 stepsv1.0.0

Deploy Brand Voice Extractor to your jetty.io

One-click installs this runbook into a collection on your Jetty account. You can run it from the Spot dashboard, schedule it, or pipe inputs in via the API.

Deploy on jetty.io →View source
The shape of the run

7 steps · start to finish.

  1. 1
    Step 1

    Environment Setup

    ▶
    1. Create results_dir if it does not exist.
    2. Confirm company_name is present.
    3. Confirm either content_urls or content_inventory_path is present.
    4. Initialize /app/results/validation_report.json with a setup stage.
  2. 2
    Step 2

    Select Content to Analyze

    ▶

    If direct URLs are provided, normalize and deduplicate them. Otherwise, read the content inventory and select a diverse sample of 10-20 pages:

  3. 3
    Step 3

    Fetch and Extract Text

    ▶

    For each selected URL:

  4. 4
    Step 4

    Analyze Voice Dimensions

    ▶

    Evaluate the sample across these dimensions and keep short evidence snippets for each finding:

  5. 5
    Step 5

    Generate Brand Voice Profile

    ▶

    Write `/app/results/brand_voice_profile.md` using this structure:

  6. 6
    Step 6

    Validate Outputs

    ▶

    Check that:

  7. 7
    Step 7

    Iterate on Errors (max 3 rounds)

    ▶

    If validation fails or the evidence is thin, run up to max 3 rounds of targeted fixes. Replace failed URLs, add missing dimensions, tighten unsupported claims, or regenerate malformed files. Stop early once all required outputs pass validation.