Tabitha Fanny
Community@tabithafanny
I'm a digital enthusiast who loves developing apps. Feel free to reach out and connect with me.
Agent Skills by Tabitha Fanny
Showing 69 vetted skills indexed across 1 GitHub repositories.
proof-checker
Verify LaTeX mathematical proofs, identify gaps, and generate audit reports.
paper-write
Generate LaTeX paper sections from user-provided outlines.
paper-poster
Convert LaTeX and TeX code into A0/A1 PDF and PPTX conference posters.
writing-systems-papers
Generate structured outlines for systems papers targeting OSDI, SOSP, ASPLOS, NSDI, and EuroSys.
qzcli
Manage GPU compute jobs on the Qizhi platform via command line.
specification-writing
Generate a patent specification from claims and invention disclosure.
proof-writer
Generate and validate mathematical proofs for machine learning theorems in Markdown.
paper-claim-audit
Verify research paper claims against raw result files with a fresh reviewer.
serverless-modal
Deploy and execute GPU-intensive tasks on Modal's serverless platform.
formula-derivation
Organize assumptions, normalize notation, and create structured derivation maps.
training-check
Monitor WandB training metrics for NaN values, loss divergence, and idle GPUs.
meta-optimize
Analyze ARIS usage logs and propose optimizations for harness components.
patent-novelty-check
Compare invention claims to prior art for novelty and non-obviousness.
vast-gpu
Manage GPU instance lifecycle on vast.ai with dynamic allocation and cost optimization.
result-to-claim
Evaluate experimental results against intended claims using Codex MCP and route next actions.
figure-spec
Convert structured JSON input into deterministic SVG architecture and workflow diagrams.
mermaid-diagram
Generate Mermaid diagrams from user descriptions and verify syntax with Mermaid CLI.
semantic-scholar
Search published papers and retrieve citation data via the Semantic Scholar API.
research-wiki
Manage and query a persistent structured knowledge base for research papers, ideas, experiments, and claims.
system-profile
Profile and analyze system performance to identify bottlenecks and generate optimization recommendations.
research-refine-pipeline
Chain research refinement and experiment planning into a proposal and roadmap.
jurisdiction-format
Format patent applications for CNIPA, USPTO, and EPO filings.
experiment-audit
Verify experimental result integrity by detecting fraud patterns with cross-model review.
auto-review-loop-llm
Automate iterative research paper reviews using OpenAI-compatible LLM APIs.