voronoi

Skill for using Beads (bd) as the single source of truth for task tracking, dependency management, and agent coordination. Covers all bd commands needed for multi-agent orchestration.

Skill for working with artifact contracts — the file-level dependency system that prevents pipeline failures by ensuring tasks produce expected outputs and only start when required inputs exist.

Skill for generating scientific figures from experimental data. Ensures every figure referenced in LaTeX documents exists on disk before compilation, preventing blank PDFs and broken references.

Complete recipe for dispatching, monitoring, merging, and cleaning up worker agents. Read this skill BEFORE dispatching any worker. Covers build_worker_prompt, spawn-agent.sh, merge-agent.sh, checkpoint updates, and common failure modes.

Use when asked to reduce code complexity, simplify code, refactor safely, merge micro-files, remove dead code, shrink functions, reduce coupling, clean modules, or improve maintainability in src/voronoi/.

Use for adversarial review of Voronoi plans, findings, claims, manuscripts, and final deliverables. Covers reviewer-style objections, confounds, alternative explanations, evidence gaps, and falsification-oriented critique.

Skill for running standup meetings across all active agents in the swarm. Aggregates Beads task status, git branch activity, commit history, and conflict detection into a human-readable progress report.

Skill for safely merging completed agent branches back to main. Covers pre-merge checks, conflict resolution, post-merge cleanup, and promotion of newly unblocked tasks.

Skill for compiling LaTeX papers with hard dependency checking. Ensures all figures, bibliography, and data dependencies are resolved before compilation, and verifies the output PDF is complete.

Skill for context window management in long-running agent sessions. Prevents context exhaustion through proactive compression and delegation. Used by orchestrators and long-running worker agents.

Skill for making programmatic LLM calls via Copilot CLI. Covers correct invocation patterns, model configuration, caching, and common error recovery. Use when any agent or script needs to call an LLM programmatically.

Skill for ensuring data integrity across investigation workflows. Covers SHA-256 hashing protocol, raw data preservation, provenance tracking, and data-file validation. Use when creating, moving, or validating experimental data files.

Use for debugging failed Voronoi worker tasks, broken verification loops, failing scripts, unexpected experiment outputs, and reproducible runtime errors. Emphasizes evidence-first reproduction before fixes.

Skill for deep research grounding using Copilot CLI /research for multi-source literature review, prior-art search, and SOTA anchoring. Use before forming hypotheses to anchor claims in real, citation-backed evidence.

Skill for managing the evidence knowledge store: creating and querying findings, maintaining belief maps and investigation journals, enforcing data integrity, and running consistency checks across validated findings.

Skill for creating, managing, and cleaning up git worktrees used as isolated working directories for parallel agent execution. Covers worktree lifecycle from creation through merge and teardown.

Skill for running the hypothesis → experiment → finding workflow in scientific investigations. Covers pre-registration, sensitivity analysis, data integrity, review gates, replication, and serendipity handling.

Use for inspecting Voronoi investigation state, event logs, progress files, tmux/worker status, .swarm artifacts, and Beads task traces without changing investigation behavior.

Skill for iterative experiment design using REVISE tasks and calibration workflows. Use when a pilot experiment fails calibration or results are unexpected and need systematic revision with full context carry-forward.

Use for creating small Voronoi investigation scaffolds, task skeletons, experiment harness placeholders, manuscript state directories, and reusable artifact templates.

Skill for maintaining the Strategic Context Document — a living document that preserves the original intent, decision rationale, dead ends, and remaining gaps across OODA cycles and sessions. Prevents context degradation in long-running tasks.

Skill for decomposing high-level feature requests into structured epics and subtasks using Beads (bd). Covers dependency graph creation, priority assignment, scope definition, and agent dispatch planning.

Use for navigating Voronoi runtime APIs available inside investigation workspaces, including prompt helpers, science validators, citation coverage, manifest checks, and utility modules.

Use when brainstorming Voronoi design, reviewing architecture, or proposing changes. Contains load-bearing decisions, rejected alternatives, known weak spots, and cross-cutting constraints. Prevents re-proposing ideas that were already tried and rejected.

Use when writing, updating, or running tests for voronoi. Use when any source code in src/voronoi/ is modified to ensure tests are updated. Covers test discovery, naming conventions, and verification workflow.

Use when brainstorming Voronoi UX from a real scientist's perspective. Contains domain knowledge about how researchers actually work, common pain points in scientific tooling, and UX patterns that respect the scientific workflow. Loaded by the Catalyst agent.