talont-org/vis-lens-story-arc

Narrative Story Arc Visualization Lens

Philosophical Mode: Narrative Primary Question: "Do the figures tell a coherent story across the report?" Focus: Visual consistency (same color = same model everywhere), logical figure progression (each figure builds on the previous), no redundant figures (same data shown twice), narrative dependency between figures (figure N motivates figure N+1)

Arguments

/autoskillit:vis-lens-story-arc [context_path] [experiment_plan_path]

• context_path (optional positional arg 1) — Absolute path to a lens context file containing IV/DV tables, H0/H1 hypotheses, controlled variables, and success criteria. If provided, read this file before beginning analysis to obtain structured context. If omitted, discover context by exploring the CWD.
• experiment_plan_path (optional positional arg 2) — Absolute path to the full experiment plan. If provided, read for complete experimental methodology and design. If omitted, locate the experiment plan by exploring the CWD.

When to Use

• Planning the figure sequence for a paper or technical report
• Checking that the same model/condition uses the same color across all figures
• Identifying redundant figures that show the same data twice
• Verifying that each figure's narrative position is justified
• User invokes /autoskillit:vis-lens-story-arc

Critical Constraints

NEVER:

• Modify any source code files
• Do not litter the codebase with useless comments, TODO markers, or explanatory annotations — the skill output and diagram speak for themselves
• Create files outside {{AUTOSKILLIT_TEMP}}/vis-lens-story-arc/
• Assign the same color to two different models, conditions, or categories across figures
• Include a figure that presents the same data and conclusion as another figure already in the plan

ALWAYS:

• Build a global color→entity mapping table across all figures; flag any inconsistency

• Number all figures and write a one-sentence narrative role for each

• Flag any figure whose narrative role duplicates another (same question, same data)

• Verify that figures appear in a logical dependency order (motivation → method → result → implication)

• The primary diagram output is a figure-sequence flow diagram (mermaid) showing narrative dependencies

• BEFORE creating any diagram, LOAD the /autoskillit:mermaid skill using the Skill tool — this is MANDATORY

• If the Skill tool cannot be used (disable-model-invocation) or refuses this invocation, do NOT proceed with diagram creation. Abort this step and omit the diagram from output.

• Write output to {{AUTOSKILLIT_TEMP}}/vis-lens-story-arc/vis_spec_story_arc_{YYYY-MM-DD_HHMMSS}.md (relative to the current working directory)

• After writing the file, emit the structured output token as literal plain text with no markdown formatting on the token name (the adjudicator performs a regex match):

  diagram_path = /absolute/path/to/{{AUTOSKILLIT_TEMP}}/vis-lens-story-arc/vis_spec_story_arc_{...}.md
  

---

Analysis Workflow

Step 0: Parse optional arguments

If positional arg 1 (context_path) is provided and the file exists, read it to obtain IV/DV tables, H0/H1 hypotheses, controlled variables, and success criteria. If positional arg 2 (experiment_plan_path) is provided and exists, read the experiment plan for full methodology. Use this structured context as the foundation for Steps 1–4; skip the CWD exploration for these fields if the context file supplies them.

Step 1: Enumerate and Number All Figures

List all planned figures in document order. For each figure record:

• Figure ID (fig-01, fig-02, ...)
• Title / description
• What data/question it addresses
• Which section it appears in (intro, method, results, discussion, appendix)

Step 2: Build Global Color Map

Scan all figure descriptions and plotting code for color/palette assignments:

• Build table: { color_hex_or_name → entity_name }
• Check: is the same entity always the same color across all figures?
• FLAG INCONSISTENCY if entity A is blue in fig-02 but orange in fig-05

Step 3: Detect Redundant Figures

For each pair of figures, check:

• Do they display the same underlying data (same x/y variables, same conditions)?
• Do they answer the same question?
• If YES → flag as REDUNDANT; recommend merging or removing one

Step 4: Map Narrative Dependencies

For each figure, identify:

• Which figure(s) it logically depends on (must be read first)
• Which figure(s) it motivates (enables interpretation of)
• Whether its narrative position (section) matches its dependency order

Step 5: Emit yaml:figure-spec Blocks and Sequence Diagram

For each figure, emit one yaml:figure-spec fenced block. Then LOAD /autoskillit:mermaid and create a figure-sequence flow diagram showing narrative dependencies between figures.

---

Output Template

# Narrative Story Arc Spec: {System / Experiment Name}

**Lens:** Narrative Story Arc (Narrative)
**Question:** Do the figures tell a coherent story across the report?
**Date:** {YYYY-MM-DD}
**Scope:** {What was analyzed}

## Global Color Map

| Color | Entity | Consistent Across Figures |
|-------|--------|--------------------------|
| #1f77b4 | Model A | PASS |
| #ff7f0e | Baseline | FAIL (blue in fig-03) |

## Figure Sequence Summary

| Figure | Section | Narrative Role | Depends On | Redundant? |
|--------|---------|----------------|------------|------------|
| fig-01 | Results | Establish main result | — | No |
| fig-02 | Results | Show ablation | fig-01 | No |

## Figure Specs

```yaml
# yaml:figure-spec — canonical schema (spec_version: "1.0")
figure_id: "fig-01-main-result"
figure_title: "Model A achieves state-of-the-art on all benchmarks"
spec_version: "1.0"
chart_type: "bar"
chart_type_fallback: "table"
perceptual_justification: "Grouped bars directly compare models; color consistent with fig-02 through fig-05."
data_source: "results/main.csv"
data_mapping:
  x: "benchmark"
  y: "score"
  color: "model"
  size: ""
  facet: ""
layout:
  width_inches: 6.0
  height_inches: 4.0
  dpi: 300
stat_overlay:
  type: "error_bar"
  measure: "CI95"
  n_seeds: 5
annotations: ["Narrative role: introduce main result; motivates fig-02 ablation"]
anti_patterns: []
palette: "okabe-ito"
format: "pdf"
target_dpi: 300
library: "matplotlib"
report_section: "Section 4 Results"
priority: "P0"
placement_tier: "main"
conflicts: []
metadata:
  created_by: "vis-lens-story-arc"
  reviewed_by: ""
  last_updated: "{YYYY-MM-DD}"

Figure Sequence Flow Diagram

%%{init: {'flowchart': {'nodeSpacing': 50, 'rankSpacing': 60, 'curve': 'basis'}}}%%
flowchart TB
    %% CLASS DEFINITIONS %%
    classDef cli fill:#1a237e,stroke:#7986cb,stroke-width:2px,color:#fff;
    classDef stateNode fill:#004d40,stroke:#4db6ac,stroke-width:2px,color:#fff;
    classDef handler fill:#e65100,stroke:#ffb74d,stroke-width:2px,color:#fff;
    classDef output fill:#00695c,stroke:#4db6ac,stroke-width:2px,color:#fff;
    classDef detector fill:#b71c1c,stroke:#ef5350,stroke-width:2px,color:#fff;

    F1["fig-01: Main Result<br/>━━━━━━━━━━<br/>Establish benchmark SOTA"]
    F2["fig-02: Ablation<br/>━━━━━━━━━━<br/>Component contribution"]
    F3["fig-03: Scaling<br/>━━━━━━━━━━<br/>Performance vs. size"]
    F4["fig-04: Error Analysis<br/>━━━━━━━━━━<br/>Failure mode taxonomy"]

    F1 --> F2
    F1 --> F3
    F2 --> F4

    class F1 stateNode;
    class F2,F3 handler;
    class F4 output;

Color Legend: | Color | Category | Description | |-------|----------|-------------| | Dark Teal | Anchor | Primary result figure | | Orange | Derived | Figures that build on anchor | | Dark Teal (output) | Terminal | Figures that conclude a narrative thread |

---

## Pre-Diagram Checklist

Before creating the diagram, verify:

- [ ] LOADED `/autoskillit:mermaid` skill using the Skill tool
- [ ] Using ONLY classDef styles from the mermaid skill (no invented colors)
- [ ] Diagram will include a color legend table
- [ ] Global color map table is complete before creating the diagram
- [ ] Every color inconsistency has been flagged
- [ ] Every redundant figure pair has been identified