als-apg/demo-gallery

Demo Gallery — Artifact Showcase Generator

Generate a diverse set of high-quality artifacts to demonstrate every capability of the OSPREY Artifact Gallery. The artifacts should be visually striking, scientifically themed, and exercise all rendering paths.

Follow these two phases in order.

---

Phase 1 — Generate Artifacts (in parallel)

Launch all four artifact groups below in parallel. Each group uses a different creation pathway to showcase the gallery's full range.

1a. Interactive Plotly Figure — create_interactive_plot

Use the create_interactive_plot MCP tool to produce a multi-panel Plotly figure with at least two subplots. The plot should be visually rich and interactive.

Suggested content (pick one or combine creatively):

• Damped harmonic oscillator with envelope + phase-space portrait
• Synchrotron radiation spectrum (intensity vs photon energy) + cumulative flux
• Particle beam phase space (x-x', y-y') with color-coded amplitude
• Resonance diagram with tune footprint overlay

Use plotly.subplots.make_subplots for multi-panel layout. Apply a clean template (plotly_white or plotly_dark). Add descriptive axis labels and a figure title. Call save_artifact(fig, "Title") at the end.

1b. Matplotlib Static Plot — execute

Use the execute MCP tool to create a publication-quality matplotlib figure and save it as an artifact. The figure should demonstrate matplotlib's strengths (colormaps, contours, annotations).

Suggested content (pick one):

• 2D heatmap of a beam distribution in phase space with contour overlay
• Twiss parameter evolution ($\beta_x$, $\beta_y$, $\eta_x$) along a beamline
• Frequency map analysis colored by diffusion rate
• Mountain range plot of bunch profiles over multiple turns

Use plt.subplots(), apply plt.style.use("seaborn-v0_8-whitegrid") or similar, and call save_artifact(fig, "Title", "description").

1c. Markdown Report with LaTeX — artifact_save

Use the artifact_save MCP tool to create a rich markdown document that exercises KaTeX rendering and table formatting.

The report MUST include:

1. A title and introduction with inline math (e.g., $E = \gamma m_0 c^2$)
2. A data table with numeric values (at least 4 rows, 4 columns)
3. Display equations using $$...$$ blocks — at least two, such as:
   • Hill's equation: $x''(s) + K(s), x(s) = 0$
   • Beam emittance: $\epsilon = \sqrt{\langle x^2 \rangle \langle x'^2 \rangle - \langle x x' \rangle^2}$
   • Synchrotron radiation critical energy: $E_c = \frac{3}{2} \frac{\hbar c \gamma^3}{\rho}$
4. Multiple heading levels (H1, H2, H3)
5. A bulleted list summarizing key observations

Use content_type: "markdown". Give it an informative title like "Accelerator Physics Reference Card" or "Beam Dynamics Summary".

1d. Computed Data Table — execute

Use the execute MCP tool to compute a numeric result and save a markdown table as an artifact.

Suggested content:

• Generate a lattice element table with columns: Element, Type, Length [m], K1 [1/m²], Angle [mrad]
• Compute a parameter scan (e.g., tune vs. quadrupole strength) and present it as a table
• Calculate beam parameters at different energies and tabulate them

Use save_artifact(summary_string, "Title", "description") where the string is a formatted markdown table. Include descriptive column headers and realistic numeric values.

---

Phase 2 — Focus & Confirm

After all artifacts are created:

1. Focus the most visually striking artifact (usually the Plotly figure) using artifact_focus
2. List all created artifacts in a summary table for the user:

| # | Title | Type | Pathway | |---|-------|------|---------| | 1 | ... | Interactive plot | create_interactive_plot | | 2 | ... | Static plot | execute + matplotlib | | 3 | ... | Markdown report | artifact_save | | 4 | ... | Data table | execute + save_artifact() |

3. Note that additional notebook artifacts were auto-generated from each execute call

---

Content Guidelines

• Domain: Use accelerator physics, beam dynamics, or synchrotron science content. This fits the OSPREY context and produces visually interesting results.
• Data: Generate synthetic data with numpy. Use realistic parameter ranges (e.g., beam energies 1-8 GeV, tunes 10-30, beta functions 1-30 m).
• Quality: Plots should look publication-ready. Use proper axis labels, units, legends, and colorbars where appropriate.
• Variety: Each artifact should look distinctly different — vary color schemes, plot types, and content.
• LaTeX: Use proper accelerator physics notation. The gallery renders LaTeX via KaTeX — use $...$ for inline and $$...$$ for display math. Inside Python string literals in execute, escape backslashes (\\\\) for LaTeX commands.

Anti-Patterns

Do NOT:

• Create all artifacts sequentially — Phase 1 groups must run in parallel
• Use trivial or placeholder content — make it visually impressive for demos
• Skip the LaTeX in the markdown report — math rendering is a key gallery feature
• Create more than 4-5 artifacts total (excluding auto-notebooks) — keep it focused
• Use the same color palette across all plots — vary the visual style
• Skip artifact_focus — the demo should open to the most striking visual