UniverLab
Official@UniverLab · Colombia
SCI · CLI · BIO
Agent Skills by UniverLab
Showing 10 vetted skills indexed across 1 GitHub repositories.
UniverLabexecution-mindset
Enforces interview-driven alignment, systematic problem resolution, and verification before reporting in agent workflows.
UniverLabcanopy-sync
Coordinates shared workspace actions through a risk-classified sync protocol for Canopy MCP sessions.
UniverLabdemostage
Captures terminal sessions and compiles them into reproducible gif and mp4 demos.
UniverLabcanopy-intelligence
Pulls workspace context and persists durable project knowledge via Canopy MCP intelligence tools.
UniverLabcanopy-capabilities
Maps Canopy workspace capabilities to user needs and discovers the live tool surface.
UniverLabcadspec
Generates and validates declarative CAD geometry in TOML files compiled to DXF drawings.
UniverLabtexforge
Compile LaTeX documents to PDF using tectonic without external TeX distributions.
UniverLabarchitect-mindset
Guides system design decisions through boundaries, contracts, tradeoffs, and executable specs.
UniverLabcode-engineering
Guides implementation, refactoring, and review of production code using structural design heuristics.
UniverLabcanopy-loop-design
Designs reusable Canopy loop graphs with agent, check, and gate nodes via MCP tools.
Frequently Asked Questions About UniverLab
FAQPage SchemaWhat tasks can I accomplish with UniverLab's skills?▼
You can coordinate multi-agent Canopy workspaces via sync and intelligence protocols, design reusable loop graphs with approvals and retries, compile LaTeX to PDF without TeX Live, build declarative CAD drawings in TOML exported to DXF, and record terminal demos as reproducible demo.toml scores exported to GIF or MP4.
Who are UniverLab's skills designed for?▼
They target engineers and technical authors working in OpenCode or Canopy MCP environments: developers needing disciplined execution and code-review judgment, system designers writing specs and decision records, CAD practitioners adopting CAD-as-code, and academics compiling theses or papers without a local LaTeX distribution.
What are the prerequisites for using these skills?▼
The four canopy-* skills require an available Canopy MCP server in the session. demostage, cadspec, and texforge run under the OpenCode framework as CLI-driven skills. execution-mindset, architect-mindset, and code-engineering are behavior skills with no external dependencies.
Are UniverLab's skills free and open source?▼
Yes. All ten skills in the manifest are released under the MIT license, permitting free use, modification, and redistribution. They are authored by jheison.martinez and published through the UniverLab GitHub organization with no stated cost or usage fees.
How do the Canopy skills work together in practice?▼
canopy-capabilities maps what the workspace can do, canopy-sync governs safe writes and commits via the action-driven sync protocol, canopy-intelligence persists facts and patterns through the Project Intelligence Layer, and canopy-loop-design turns recurring processes into reusable loop graphs with gates, approvals, and retries.