EvoClaw
Official@evoclaw
Offers rigorous research governance, manuscript validation, and reproducible experimental orchestration for scientific publication cycles.
Agent Skills by EvoClaw
Showing 24 vetted skills indexed across 1 GitHub repositories.
EvoClawfigure-quality-standards
Enforce publication-quality standards for figures and tables in scientific papers.
EvoClawnovelty-classifier
Classify research contribution novelty and verify target venue tier alignment.
EvoClawclaim-evidence-alignment
Validate manuscript claims against linked figures, tables, experiments, and statistical results.
EvoClawvenue-alignment
Enforce project progress alignment with target venue tier requirements at gate reviews.
EvoClawalternative-hypothesis-check
Identifies and rules out alternative explanations for study findings.
EvoClawscope-control
Quantify scope creep risk and require user approval before applying project changes.
EvoClawanti-cherry-pick
Enforce complete reporting of all seeds, results, and failures in experiments.
EvoClawdispatching-parallel-agents
Dispatch parallel subagents to run independent experiments concurrently.
EvoClawresults-verification-protocol
Run fresh verification commands and validate research claims against output evidence.
EvoClawpaper-writing
Transform research results into a publication-ready LaTeX manuscript with verified references.
EvoClawexperiment-execution
Orchestrate Phase 4 experimental execution with fixed seeds and environment snapshots.
EvoClawpivot-or-kill
Escalate research decisions after three consecutive core experiment failures.
EvoClawresearch-direction-exploration
Identify viable research directions from vague inputs and literature signals.
EvoClawevaluation-protocol-design
Define metrics, datasets, seeds, baselines, and fairness constraints for evaluation protocols.
EvoClawmulti-round-deliberation
Coordinate iterative multi-agent discussions to converge on a resolved artifact.
EvoClawmethod-framework-design
Generate Phase 3 research design packages with evaluation protocols and baselines.
EvoClawanalysis-storyboard-design
Plan analysis storyboards with main and supporting lines for discovery workflows.
EvoClawproblem-validation
Orchestrate multi-agent adversarial validation of research questions for novelty, feasibility, and venue fit.
EvoClawresults-integration
Transform research results into an evidence-backed manuscript outline and narrative structure.
EvoClawdomain-anchoring
Identify research domain and subdomain, classify research type, and generate research-anchor.yaml.
EvoClawusing-amplify
Enforce one-phase-per-turn research workflows with mandatory gates G1 through G4.
EvoClawreproducibility-driven-research
Enforce reproducible research cycles with hypothesis-first planning and scripted experiments.
EvoClawmetric-lock
Lock evaluation metrics, protocols, datasets, and seeds after plan freeze.
EvoClawusing-git-worktrees
Create isolated git worktrees with safety checks before switching branches.
Frequently Asked Questions About EvoClaw
FAQPage SchemaWhat specific research tasks does EvoClaw enable?▼
EvoClaw enables rigorous research governance, including novelty classification, claim-to-evidence validation, and the transformation of experimental results into publication-ready LaTeX manuscripts. It enforces strict gate reviews and reproducibility standards throughout the research lifecycle.
Which personas benefit from these research capabilities?▼
Academic researchers, data scientists, and laboratory leads benefit from these capabilities. The system is designed for professionals requiring strict adherence to scientific rigor, reproducible experimental design, and systematic manuscript preparation for high-tier publication venues.
What are the prerequisites for running EvoClaw research cycles?▼
Users require a defined research hypothesis, target venue requirements, and access to a computational environment capable of supporting Git worktrees. The system mandates a structured, phase-based approach with fixed seeds and environment snapshots to ensure full reproducibility.