enhance-glyph
À propos
Cette compétence améliore les glyphes pictogrammes existants basés sur R en diagnostiquant et en corrigeant les problèmes visuels tels que les mauvaises proportions, les problèmes de lisibilité ou les effets de lueur déséquilibrés. Elle guide les développeurs à travers l'audit du glyphe, la modification ciblée des fonctions et la comparaison des résultats avant/après. Utilisez-la lorsque les glyphes s'affichent mal à petite taille, ont des métaphores visuelles peu claires, ou nécessitent des mises à jour après des changements de palette ou de pipeline.
Installation rapide
Claude Code
Recommandénpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/enhance-glyphCopiez et collez cette commande dans Claude Code pour installer cette compétence
Documentation
Enhance Glyph
Improve existing pictogram glyph in viz/ viz layer — audit how it renders, diagnose issues, apply tight fixes, re-render, and compare before/after. Works for skill, agent, team glyphs.
When Use
- Glyph renders poor at small size (details lost, shapes merge)
- Glyph visual meaning unclear or not match entity
- Glyph has proportion issue (too big, too small, off-center)
- Neon glow eats glyph or too weak
- Glyph good in one palette, poor in others
- Batch fix after new palette or new render pipeline
Inputs
- Required: Entity type —
skill,agent, orteam - Required: Entity ID of glyph to enhance (e.g.,
commit-changes,mystic,tending) - Required: Specific issue to fix (readability, proportions, glow, palette compat)
- Optional: Reference glyph that shows target quality
- Optional: Target palette(s) to tune for (default: all palettes)
Steps
Step 1: Audit — Check Current State
Look at current glyph; spot specific issues.
- Find glyph function by entity type:
- Skills:
viz/R/primitives*.R(19 domain-grouped files), mapped inviz/R/glyphs.R - Agents:
viz/R/agent_primitives.R, mapped inviz/R/agent_glyphs.R - Teams:
viz/R/team_primitives.R, mapped inviz/R/team_glyphs.R
- Skills:
- Read glyph function to grasp shape:
- How many layers?
- What primitives call?
- What scale factors and place?
- View rendered output:
- Skills:
viz/public/icons/cyberpunk/<domain>/<skillId>.webp - Agents:
viz/public/icons/cyberpunk/agents/<agentId>.webp - Teams:
viz/public/icons/cyberpunk/teams/<teamId>.webp - If can, check 2-3 other palettes for cross-palette render
- View at icon size (~48px in graph) and panel size (~160px in detail panel)
- Skills:
- Score glyph on quality dimensions:
Glyph Quality Dimensions:
+----------------+------+-----------------------------------------------+
| Dimension | 1-5 | Assessment Criteria |
+----------------+------+-----------------------------------------------+
| Readability | | Recognizable at 48px? Clear at 160px? |
| Proportions | | Well-centered? Good use of the 100x100 canvas?|
| Metaphor | | Does the shape clearly represent the entity? |
| Glow balance | | Glow enhances without overwhelming? |
| Palette compat | | Looks good across cyberpunk + viridis palettes?|
| Complexity | | Appropriate layer count (not too busy/sparse)? |
+----------------+------+-----------------------------------------------+
- Spot 1-2 dimensions with lowest scores — these are fix targets
Got: Clear pick of what wrong with glyph and which dimension to fix. Audit specific: "proportions: glyph uses only 40% of canvas" not "looks bad."
If fail: Glyph function missing or entity not in its *_glyphs.R map? Glyph maybe not made yet — use create-glyph instead.
Step 2: Diagnose — Root Cause
Find why issues exist.
- For readability issues:
- Too many fine details that merge at small size?
- Weak contrast between glyph elements?
- Lines too thin (< 1.5
sizeat s=1.0)? - Elements too close?
- For proportion issues:
- Scale factor
stoo small or too big? - Center off from (50, 50)?
- Elements past safe area (10-90 range)?
- Scale factor
- For glow issues:
- Glyph stroke width mixes with
ggfx::with_outer_glow():- Thin lines: glow make fuzzy
- Thick fills: glow add bloom
- Many overlap elements: stacked glow make hot spots
- Glyph stroke width mixes with
- For palette compat issues:
- Glyph use hardcoded colors, not
col/brightparams? - Low-contrast palettes (cividis, mako) hide glyph?
- Glyph needs color variation some palettes miss?
- Glyph use hardcoded colors, not
- Write specific root cause for each issue
Got: Root causes that point to code change. "Glyph too small" -> "scale factor 0.6 but should be 0.8." "Glow eats" -> "three overlap filled polygons each make glow."
If fail: Root cause not clear from code read? Render glyph alone with different params to isolate. Use render_glyph() with one glyph to test.
Step 3: Modify — Apply Tight Fixes
Edit glyph function to fix diagnosed issues.
- Open file with glyph function
- Apply fixes matched to diagnosis:
- Scale/proportion: Tune
smultiplier or element offsets - Readability: Simplify complex parts, thicker stroke, add space
- Glow balance: Fewer overlap filled areas, use outlines where fills bloom
- Palette compat: All colors from
col/brightparams, add alpha for depth
- Scale/proportion: Tune
- Follow glyph function contract:
glyph_name <- function(cx, cy, s, col, bright) { # cx, cy = center (50, 50) # s = scale (1.0 = ~70% of canvas) # col = domain color, bright = brightened variant # Returns: list() of ggplot2 layers } - Keep function signature — do not change params
- Keep changes tight: fix diagnosed issues, do not redesign whole glyph
Got: Modified glyph function that fix specific issues from Steps 1-2. Changes tight and minimal — enhance, not redesign.
If fail: Fixes make other dimensions worse (e.g., fixing proportions breaks readability)? Revert and try other path. If glyph need full redesign, use create-glyph instead.
Step 4: Re-render — Make New Icons
Render modified glyph and check fix. Always use build.sh — it handles platform detect and R binary pick. See render-icon-pipeline for full flag list.
-
Re-render by entity type:
# From project root — use --no-cache to force re-render of modified glyph bash viz/build.sh --only <domain> --no-cache # skills bash viz/build.sh --type agent --only <id> --no-cache # agents bash viz/build.sh --type team --only <id> --no-cache # teams -
Check output files at expected path for each palette
-
Check file sizes — icons should be 2-15 KB (WebP):
- Under 2 KB: glyph maybe too simple or render fail
- Over 15 KB: glyph maybe too complex (too many layers)
Got: Fresh icon files made for all palettes. File sizes in expected range.
If fail: Build script errors? Check R console output for specific error. Common: missing close paren in glyph function, ref undefined primitives, or return non-list from function. If render OK but output blank, glyph layers maybe outside canvas bounds.
Step 5: Compare — Before/After Check
Check fix improved target dimensions.
- Compare old and new renderings:
- View cyberpunk palette at icon (48px) and panel (160px) sizes
- View at least 2 other palettes (one light like turbo, one dark like mako)
- Re-score quality dimensions from Step 1:
- Target dimensions should improve by at least 1 point
- Non-target dimensions should not drop
- If glyph used in force-graph, test there:
- Start HTTP server:
python3 -m http.server 8080fromviz/ - Load graph and find entity node
- Check icon renders right at default zoom and zoomed in
- Start HTTP server:
- Write what changes made and gain reached
Got: Measurable gain on target dimensions with no drop on others. Glyph looks better at both sizes and across palettes.
If fail: Gain tiny or drop happens? Revert changes and rethink diagnosis. Sometimes old glyph limits come from metaphor, not code — then metaphor itself may need change (escalate to create-glyph).
Validation Checklist
- Current glyph audited with specific issue call
- Root cause found for each issue
- Changes tight to diagnosed issues (not over-edited)
- Glyph function contract kept (signature unchanged)
- Icons re-rendered for all palettes
- Before/after compare shows gain on target dimensions
- No drop on non-target dimensions
- File sizes in expected range (2-15 KB WebP)
- Glyph renders right in force-graph context (if used)
Pitfalls
- Over-enhancement: Fix one issue then tweak everything else. Stick to diagnosed issues
- Break contract: Change function signature break render pipeline. 5-param contract is fixed
- Palette-specific tune: Make glyph perfect for cyberpunk but poor for viridis. Always check 3+ palettes
- Ignore small-size render: Pretty 160px icon that become blob at 48px is fail
- Forget re-render: Edit function without run build cmd means changes not visible
- Wrong build cmd: Skills use
build-icons.R, agents usebuild-agent-icons.R, teams usebuild-team-icons.R
See Also
- create-glyph — make new glyph from scratch (when enhance not enough)
- audit-icon-pipeline — spot which glyphs need fix across pipeline
- render-icon-pipeline — run full render pipeline after fixes
- ornament-style-mono — visual design rules that apply to glyph composition
- chrysopoeia — value extraction method parallels glyph tune (amplify gold, drop dross)
Dépôt GitHub
Compétences associées
content-collections
MétaCette compétence propose une configuration éprouvée en production pour Content Collections, un outil axé sur TypeScript qui transforme des fichiers Markdown/MDX en collections de données typées de manière sûre avec une validation Zod. Utilisez-la lors de la création de blogs, de sites de documentation ou d'applications Vite + React riches en contenu pour garantir la sécurité de typage et la validation automatique du contenu. Elle couvre tout, de la configuration du plugin Vite et de la compilation MDX à l'optimisation des déploiements et la validation des schémas.
polymarket
MétaCette compétence permet aux développeurs de créer des applications avec la plateforme de marchés prédictifs Polymarket, incluant l'intégration d'API pour le trading et les données de marché. Elle fournit également une diffusion de données en temps réel via WebSocket pour surveiller les transactions en direct et l'activité du marché. Utilisez-la pour mettre en œuvre des stratégies de trading ou pour créer des outils traitant les mises à jour de marché en direct.
creating-opencode-plugins
MétaCette compétence aide les développeurs à créer des plugins OpenCode qui s'interconnectent avec plus de 25 types d'événements tels que les commandes, les fichiers et les opérations LSP. Elle fournit la structure du plugin, les spécifications de l'API événementielle et les modèles d'implémentation pour les modules JavaScript/TypeScript. Utilisez-la lorsque vous avez besoin d'intercepter, de surveiller ou d'étendre le cycle de vie de l'assistant IA OpenCode avec une logique personnalisée pilotée par les événements.
sglang
MétaSGLang est un framework de service LLM haute performance spécialisé dans la génération rapide et structurée pour les workflows JSON, regex et agentiques grâce à son cache de préfixe RadixAttention. Il offre une inférence nettement plus rapide, particulièrement pour les tâches avec des préfixes répétés, ce qui le rend idéal pour les sorties complexes et structurées ainsi que les conversations multi-tours. Choisissez SGLang plutôt que des alternatives comme vLLM lorsque vous avez besoin d'un décodage contraint ou que vous construisez des applications avec un partage étendu de préfixes.
