Dyscolab agent skills

Agent Skills that teach coding agents how to model systems with the Dyscolab libraries:

Skill	What it covers
dyscolab-modeling	Declaring, simulating and analysing models with poincare (ODEs), simbio (reaction networks, compartments, SBML) and jablonski (photophysics)

The skill on main documents poincare 1.2.x, simbio 1.2.x and jablonski 0.4.x.

Install

The skills use the open SKILL.md format, so they work with any agent that supports Agent Skills.

Any agent (Codex, Cursor, GitHub Copilot, Gemini CLI, Claude Code, Windsurf, Goose, …) with the skills CLI:

npx skills add dyscolab/skills

Claude Code, as a plugin marketplace:

claude plugin marketplace add dyscolab/skills
claude plugin install dyscolab@dyscolab

Manually: copy skills/dyscolab-modeling/ into your agent's skills directory (for example .agents/skills/, .claude/skills/ or ~/.codex/skills/).

The skill expects the libraries in the Python environment the agent uses:

pip install poincare "simbio[io]" jablonski rebop

Repository layout

skills/<name>/SKILL.md         # the skill (open Agent Skills format)
skills/<name>/references/      # detailed docs, loaded on demand
skills/<name>/examples/        # runnable scripts with physics checks
.claude-plugin/marketplace.json  # Claude Code marketplace manifest (points at skills/)
tests/                         # runs every example and validates the metadata

Development

pixi run test              # Python 3.14, pinned library versions
pixi run -e py312 test     # also py313

main only changes at releases. Unreleased work, including documentation for library features that are not on PyPI yet, goes to dev. See AGENTS.md.

License

MIT