synthetic-walkthrough-spec-eval

Grade per-persona walkthrough specification YAML files across six weighted dimensions.

1|2|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/dimagi-internal/ace --skill synthetic-walkthrough-spec-eval
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: synthetic-walkthrough-spec-eval
Source: https://github.com/dimagi-internal/ace/tree/main/skills/synthetic-walkthrough-spec-eval
Command: npx skills add https://github.com/dimagi-internal/ace --skill synthetic-walkthrough-spec-eval

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Poorly written per-persona walkthrough specs waste production time generating decks that fail to persuade target viewers like funders or prospective LLOs, missing critical wow moments, falsifiable quality checks, and persona alignment.

Core Features & Use Cases

  • Weighted dimensional scoring: Grades specs across 6 weighted dimensions including persona priority coverage, wow-moment specificity, AI quality falsifiability, out-of-chain persona resonance, anomaly-to-scene mapping, and persona turn-off avoidance.
  • Hard-gate guardrails: Blocks specs that would not persuade their target viewer (inert, no decision arc, checklist-only) or are too thin for the persona story arc, preventing wasted AVD rendering time.
  • Inflation guard: Caps scores for specs that pass all coverage and falsifiability checks but lack persuasive arc, avoiding false high ratings from box-ticking rubrics.
  • Use Case: Used in ACE Phase 7 Stage 4 to grade specs emitted by the synthetic-walkthrough-spec skill before passing them to the canopy:walkthrough workflow for deck generation.

Quick Start

Use the synthetic-walkthrough-spec-eval skill to evaluate the generated per-persona walkthrough spec for your current ACE opportunity and output a structured verdict YAML file.

Frequently Asked Questions about synthetic-walkthrough-spec-eval

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I evaluate a walkthrough spec for persuasive deck quality before rendering?

Spec grading checks falsifiable AI quality assertions, scene coverage, and persona alignment within walkthrough specification YAML files. This process verifies that synthetic walkthrough specs contain persuasive decision arcs rather than functioning as simple checklists before deck generation.

How do I grade persona alignment in a synthetic walkthrough specification?

Weighted dimensional scoring grades persona alignment across six dimensions including persona priority coverage, out-of-chain resonance, and turn-off avoidance. This prevents false high ratings from coverage-complete but non-persuasive box-ticking rubrics.

Can I block non-persuasive walkthrough specs from advancing to deck generation?

Yes, hard-gate guardrails block non-persuasive specs from advancing by triggering failures for inert content, missing decision arcs, or checklist-only formats. This prevents wasted production time generating decks that fail to persuade target viewers.

What happens if a walkthrough spec passes all coverage checks but lacks a persuasive arc?

An inflation guard caps the score for specs that pass all coverage and falsifiability checks but lack a persuasive arc. This prevents false high ratings from coverage-complete but non-persuasive box-ticking rubrics, ensuring target viewers are actually persuaded.

When do I need walkthrough spec grading in the ACE Phase 7 workflow?

Walkthrough spec grading is needed during Stage 4 of ACE Phase 7 Plan B to evaluate specs before they are consumed by the canopy walkthrough deck generation workflow. This ensures per-persona specs contain critical wow moments and falsifiable quality checks.