extract-story-intent

Steps:

1. Read the Provided Document/Text

• Read the provided document or text. You will be extracting information from this document to create the user story JSON.

2. Create JSON File

• create a new json file in the same directory as the document or text, the {Verb Object Context} of the user story as the filename. e.g. "create-object-with-validation.intent.json"
  • IF unable to write a file, put the extracted intent JSON in the Chat.
• The json MUST follow /schema/story-intent.schema.json.
• set the raw_request to the file path, relative to repo root, of the provided document, or to the verbatim text provided if no document was sent.

3. Extract User Story

Extract: SO THAT [benefit] → Why? What value does this provide? AS A [role] → Who is the user? What's their context? I WANT [capability] → What feature/behavior is requested?

Fill in the appropriate section of the json with this information

4. Extract Target Persona

Document: USER PROFILE: ├── Persona Name: [e.g., Sarah - E-commerce Shopper] ├── Role: [e.g., Customer, Admin, Guest] ├── Technical level: [Novice/Intermediate/Expert] ├── Frequency of use: [Daily/Weekly/Occasional] ├── Environment: [Desktop/Mobile/Both] └── Permissions: [What can they access?]

Questions to answer:

• What other actions does this user typically perform?
• What's their journey before/after this feature?
• Are there different user types affected?

Fill in the appropriate section of the json with this information

5. Extract Capability Breakdown

Break down what's being requested:

CAPABILITY ANALYSIS:
├── Core action: [The main thing user does]
├── Inputs required: [What data is needed?]
├── Outputs expected: [What should result?]
├── State changes: [What data is modified?]
└── Side effects: [Notifications, logs, etc.]

Fill in the appropriate section of the JSON. Leave the affected components empty for now.

6. Extract the Listed Acceptance Criteria

For each extract:
- verbatim criteria
- gherkin/scenario expression
- type {functional, non-functional, business rule, constraint}
- job to be done (jtbd) in Verb Object Context format if possible, ensure the JTBD is stated from the user's perspective, and value. e.g. 'Enforce System Standards' becomes something like 'Enter valid data' or 'Complete form without errors'
- persona served