mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-04-25 00:51:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
5839 commits 895 branches 10 tags 1.7 GiB
Commit graph

6 commits

Author SHA1 Message Date
SHL0MS
941608cdde feat(skills): add creative divergence strategies for experimental output 
Adds opt-in creative thinking frameworks to ascii-video, p5js, and
manim-video skills, based on Lluminate (joelsimon.net/lluminate).

Only engaged when the user explicitly asks for creative, experimental,
or unconventional output. Straightforward requests are unaffected.

Each skill gets 2-3 strategies matched to its domain:
- ascii-video: Forced Connections, Conceptual Blending, Oblique Strategies
- p5js: Conceptual Blending, SCAMPER, Distance Association
- manim-video: SCAMPER, Assumption Reversal

Strategies sourced from creativity research (Boden, Eno, de Bono,
Koestler, Fauconnier & Turner, Osborn), formalized for LLM prompting
by Lluminate.
 2026-04-09 21:40:16 -04:00
Leo Torres
e8f6854cab docs: expand Manim CE reference docs with additional API coverage 
Add geometry mobjects, movement/creation animations, and LaTeX
environments to the skill's reference docs. All verified against
Manim CE v0.20.1.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-07 11:36:13 +02:00
SHL0MS
a912cd4568 docs(manim-video): add 5 new reference files — design thinking, updaters, paper explainer, decorations, production quality 
Five new reference files expanding the skill from rendering knowledge
into production methodology:

animation-design-thinking.md (161 lines):
  When to animate vs show static, concept decomposition into visual
  beats, pacing rules, narration sync, equation reveal strategies,
  architecture diagram patterns, common design mistakes.

updaters-and-trackers.md (260 lines):
  Deep ValueTracker mental model, lambda/time-based/always_redraw
  updaters, DecimalNumber and Variable live displays, animation-based
  updaters, 4 complete practical patterns (dot tracing, live area,
  connected diagram, parameter exploration).

paper-explainer.md (255 lines):
  Full workflow for turning research papers into animations. Audience
  selection, 5-minute template, pre-code gates (narration, scene list,
  style contract), equation reveal strategies, architecture diagram
  building, results animation, domain-specific patterns for ML/physics/
  biomedical papers.

decorations.md (202 lines):
  SurroundingRectangle, BackgroundRectangle, Brace, arrows (straight,
  curved, labeled), DashedLine, Angle/RightAngle, Cross, Underline,
  color highlighting workflows, annotation lifecycle pattern.

production-quality.md (190 lines):
  Pre-code, pre-render, post-render checklists. Text overlap prevention,
  spatial layout coordinate budget, max simultaneous elements, animation
  variety audit, tempo curve, color consistency, data viz minimums.

Total skill now: 14 reference files, 2614 lines.
 2026-04-06 13:51:36 -04:00
SHL0MS
b26e7fd43a fix(manim-video): recommend monospace fonts — proportional fonts have broken kerning in Pango 
Manim's Pango text renderer produces broken kerning with proportional
fonts (Helvetica, Inter, SF Pro, Arial) at all sizes and resolutions.
Characters overlap and spacing is inconsistent. This is a fundamental
Pango limitation.

Changes:
- Recommend Menlo (monospace) as the default font for ALL text
- Proportional fonts only acceptable for large titles (>=48, short strings)
- Set minimum font_size=18 for readability
- Update all code examples to use MONO='Menlo' pattern
- Remove Inter/Helvetica/SF Pro from recommendations
 2026-04-06 00:35:43 -04:00
SHL0MS
447ec076a4 docs(manim-video): expand references with comprehensive Manim CE and 3b1b patterns 
Adds 601 lines across 6 reference files, sourced from deep review of:
- Manim CE v0.20.1 full reference manual
- 3b1b/manim example_scenes.py and source modules
- 3b1b/videos production CLAUDE.md and workflow patterns
- Manim CE thematic guides (voiceover, text, configuration)

animations.md: always_redraw, TracedPath, FadeTransform,
  TransformFromCopy, ApplyMatrix, squish_rate_func,
  ShowIncreasingSubsets, ShowPassingFlash, expanded rate functions

mobjects.md: SVGMobject, ImageMobject, Variable, BulletedList,
  DashedLine, Angle/RightAngle, boolean ops, LabeledArrow,
  t2c/t2f/t2s/t2w per-substring styling, backstroke for readability,
  apply_complex_function with prepare_for_nonlinear_transform

equations.md: substrings_to_isolate, multi-line equations,
  TransformMatchingTex with matched_keys and key_map,
  set_color_by_tex

graphs-and-data.md: Graph/DiGraph with layout algorithms,
  ArrowVectorField/StreamLines, ComplexPlane/PolarPlane

camera-and-3d.md: ZoomedScene with inset zoom,
  LinearTransformationScene for 3b1b-style linear algebra

rendering.md: manim.cfg project config, self.next_section()
  chapter markers, manim-voiceover plugin with ElevenLabs/GTTS
  integration and bookmark-based audio sync
 2026-04-06 00:08:17 -04:00
SHL0MS
9d885b266c feat(skills): add manim-video skill for mathematical and technical animations 
Production pipeline for creating 3Blue1Brown-style animated videos
using Manim Community Edition. The agent handles the full workflow:
creative planning, Python code generation, rendering, scene stitching,
audio muxing, and iterative refinement.

Modes: concept explainers, equation derivations, algorithm
visualizations, data stories, architecture diagrams, paper explainers,
3D visualizations.

9 reference files, setup verification script, README.
All API references verified against ManimCommunity/manim source.
 2026-04-05 19:09:37 -04:00