Outlines, selected nodes, components, and anchors
Use this workflow when an agent needs to inspect selected nodes, review paths, add components or anchors, or make bounded outline edits in the Glyphs file you already have open.
For recurring outline work, start with the glyphs-mcp-outlines-docs skill. It keeps the agent focused on the current font, master, glyph, layer, and selection, and it uses bundled docs lookup only when API details are uncertain.
When to use it
- You want to inspect selected nodes or current selected glyphs.
- You want to add a component or anchor to a known glyph/layer.
- You want to export path JSON for review or controlled editing.
- You want to review collinear curve handles before setting nodes to smooth.
- You need a targeted GlyphsApp API lookup through
docs_searchanddocs_get.
What changes
Only explicit mutating tools change font state:
add_component_to_glyphcan add a component.add_anchor_to_glyphcan add an anchor.set_glyph_pathscan replace paths.apply_collinear_handles_smoothcan set smooth flags after review.add_corner_to_all_masterscan add corner hints at selected nodes.
Mutating calls should be scoped to named glyphs, masters, layers, paths, or selected nodes, then verified with a read-back call.
What does not change
- The font is not auto-saved.
- Kerning and spacing do not change unless you call separate kerning or spacing tools.
- Bundled docs lookup does not mutate anything.
get_selected_nodes,get_glyph_paths, andreview_collinear_handlesare read-only.
Safe prompt template
Use the glyphs-mcp-outlines-docs skill.
Task: Inspect my current selected nodes and propose the safest next step.
Rules:
- Read current font, master, glyph, layer, and selection first.
- Prefer dedicated outline tools before execute_code.
- Use docs_search and docs_get only for the specific Glyphs API detail you need.
- Do not mutate until you show exact targets and I approve.
- After any mutation, re-read the affected glyph/layer and report changed, skipped, and unresolved items.