Arness Spark Stress Interview

Stress-test a product concept by conducting structured synthetic user interviews. This skill generates 3 synthetic personas from the product concept's persona moulds, applies adversarial casting overlays (Pragmatist, Skeptic, Power User), and runs each through a 3-phase "Two-Part Reveal" interview protocol:

1. Phase 1 -- Blind Problem Check: The persona hears about the problem without knowing the product exists. Do they even care?
2. Phase 2 -- Elevator Pitch Reveal: The full product concept is revealed. What resonates? What falls flat? What is missing?
3. Phase 3 -- Dealbreaker Probe: The weakest aspects are presented head-on. Would they still use this?

The process produces brutally honest feedback from 3 distinct adversarial perspectives, synthesized into actionable concept update recommendations. The product concept is read but never modified -- all recommendations are captured in the interview report for later review.

Prerequisites

Configuration Check

1. Read the project's CLAUDE.md and check for a ## Arness section
2. If found, extract the configured Vision directory and Reports directory paths
3. If no ## Arness section exists or Arness Spark fields are missing, inform the user: "Arness Spark is not configured for this project yet. Run /arn-brainstorming to get started — it will set everything up automatically." Do not proceed without it.
4. Create the reports directory structure if it does not exist: mkdir -p <reports-dir>/stress-tests/

Data Availability

Persona moulds fallback cascade:

If the product concept exists but the Target Personas section is missing or contains "Not explored" sentinel:

Ask (using AskUserQuestion): "The product concept does not include persona moulds, which are needed to generate synthetic interview subjects. How would you like to proceed?"

1. Run /arn-spark-discover to generate personas through product discovery
2. Describe 3 target user types now (I will generate personas from your descriptions)
3. Skip the interview stress test

If the user chooses option 2, collect brief descriptions and invoke arn-spark-persona-architect in instantiation mode to generate persona instances with casting overlays before proceeding.

Workflow

Step 1: Load References

Load the interview protocol, persona casting spec, and report template:

Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-spark-stress-interview/references/interview-protocol.md Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-spark-stress-interview/references/persona-casting-spec.md Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-spark-stress-interview/references/interview-report-template.md

Step 2: Read Product Concept and Extract Context

Read the product concept from <vision-dir>/product-concept.md. Extract:

• Product vision and problem statement (for Phase 1 blind problem check)
• Full product concept summary (for Phase 2 reveal)
• Product pillars (for targeted probing)
• Competitive landscape (for Phase 3 competitive comparisons)
• Persona moulds from the Target Personas section
• Scope boundaries and deferred features (for Phase 3 dealbreaker probes)

Step 3: Generate Synthetic Personas (Casting)

Select 3 persona moulds from the product concept. If more than 3 moulds exist, select the 3 most diverse in adoption posture and technical sophistication.

For each mould, invoke the arn-spark-persona-architect agent in instantiation mode via the Task tool, passing the model from .arness/agent-models/spark.md as the model parameter (see plugins/arn-spark/skills/arn-spark-ensure-config/references/ensure-config.md "Dispatch convention" for fallback). Context:

--- PERSONA MOULD --- [full abstracted profile from product concept] --- END PERSONA MOULD ---

--- CASTING OVERLAY --- [overlay specification from persona-casting-spec.md -- one of Pragmatist, Skeptic, Power User] --- END CASTING OVERLAY ---

--- PRODUCT CONTEXT --- [product vision and problem statement for domain grounding] --- END PRODUCT CONTEXT ---

Assign overlays following the priority in the casting spec: natural fit first, then maximize coverage.

Receive back 3 fully detailed concrete personas with casting overlays baked in. Present a brief summary to the user:

"I have generated 3 synthetic interview subjects:

• [Persona 1 Name] ([Archetype]) with Pragmatist lens -- focused on practical adoption barriers
• [Persona 2 Name] ([Archetype]) with Skeptic lens -- focused on trust and systemic risk
• [Persona 3 Name] ([Archetype]) with Power User lens -- focused on depth and scalability

Starting interviews now. Each persona goes through 3 phases: blind problem check, full reveal, and dealbreaker probe."

Step 4: Conduct Interviews (3 Phases x 3 Personas -- Phase-Parallel)

Run interviews in 3 waves. Each wave runs one phase for all 3 personas in parallel. Each persona's interview context is independent -- no cross-persona information is shared between parallel invocations.

Wave 1 -- Phase 1 (Blind Problem Check)

Run all 3 personas' Phase 1 interviews in parallel. For each persona in parallel:

Invoke the arn-spark-product-strategist agent via the Task tool, passing the model from .arness/agent-models/spark.md as the model parameter (see plugins/arn-spark/skills/arn-spark-ensure-config/references/ensure-config.md "Dispatch convention" for fallback). Context:

--- PRODUCT CONCEPT --- [full product concept] --- END PRODUCT CONCEPT ---

--- INTERVIEW PROTOCOL --- Phase 1 goal: formulate 2-3 questions that probe problem recognition without revealing the solution. Strip solution-specific language from the problem description. --- END INTERVIEW PROTOCOL ---

--- PERSONA PROFILE --- [full cast persona profile] --- END PERSONA PROFILE ---

Then invoke the arn-spark-persona-impersonator agent via the Task tool, passing the model from .arness/agent-models/spark.md as the model parameter (see plugins/arn-spark/skills/arn-spark-ensure-config/references/ensure-config.md "Dispatch convention" for fallback). Context:

--- PERSONA PROFILE --- [full cast persona profile] --- END PERSONA PROFILE ---

--- CASTING OVERLAY --- [overlay name and specification] --- END CASTING OVERLAY ---

--- PRODUCT CONCEPT SUMMARY --- [problem space description ONLY -- no product details] --- END PRODUCT CONCEPT SUMMARY ---

--- INTERVIEW PHASE --- reaction --- END INTERVIEW PHASE ---

--- INTERVIEW PROMPT --- [questions from product strategist] --- END INTERVIEW PROMPT ---

Record the full response for each persona.

After all 3 personas' Phase 1 responses are collected, proceed to Wave 2.

Wave 2 -- Phase 2 (Deep Probing)

After Wave 1 completes, run all 3 personas' Phase 2 interviews in parallel. Each persona's Phase 2 needs only its own Phase 1 context (available from Wave 1). For each persona in parallel:

Invoke arn-spark-product-strategist with that persona's Phase 1 responses and the full product concept, requesting Phase 2 questions that target gaps between the persona's described problem and the product's proposed solution.

Then invoke arn-spark-persona-impersonator with the full product concept revealed, interview phase set to probing, and questions from the strategist. Include that persona's Phase 1 responses as prior context.

Record the full response for each persona.

After all