Nishant Sharma
Community@hellonish · USA
AI Engineer, ML Researcher
Agent Skills by Nishant Sharma
Showing 24 vetted skills indexed across 1 GitHub repositories.
financial_search
Aggregate SEC EDGAR filings and financial news with credibility scoring.
patent_search
Search Google Patents and Espacenet for structured patent metadata.
video_search
Search YouTube videos and extract transcripts with channel and duration metadata.
legal_search
Retrieve jurisdiction-filtered case law and judicial opinions from CourtListener.
standards_search
Identify metadata-only references to technical standards from user queries.
data_extraction
Extract tabular data from PDFs and HTML into a structured 2D string matrix.
book_search
Query the Google Books API to retrieve structured book metadata and preview links.
news_archive
Retrieves recent news articles from approved domains with structured metadata.
knowledge_delta
Compare current findings against a prior state to produce a JSON delta.
annotation_gen
Generate per-source critical annotations for academic citations and summaries.
exec_summary
Generate a 350-word executive summary with bottom line, three evidences, limitation, and next step.
visualization_spec
Generate a declarative JSON chart specification from structured data.
explainer
Translate jargon-heavy technical content into clear, accessible explanations with analogies.
report_generator
Synthesize upstream research artifacts into a five-section report with inline citations.
decision_matrix
Generate a markdown decision matrix with condition-driven recommendations and caveats.
claim_verification
Map synthesis claims to cited sources and output per-claim verdicts.
causal_analysis
Evaluate causal relationships in data and text using evidence type taxonomy.
hypothesis_gen
Generates testable hypotheses with rationale, scores, and suggested tests from evidence gaps.
contradiction_detect
Pair claims from distinct sources to detect and characterize cross-source contradictions.
synthesis
Aggregate upstream analysis claims into a cited narrative with consensus notes.
trend_analysis
Convert dated observations into structured trend objects with deterministic JSON output.
timeline_construct
Extract dated events from upstream sources and produce a cited chronological timeline.
meta_analysis
Pool effect sizes from clinical studies into a single estimate with I².
statistical_analysis
Compute mean, median, range, and count for numeric columns in tabular data.