Research & Search

Compile deep repository evidence into artifact-bound correctness doctrine, including authority/law/proof maps, owned invariants, failure archaeology, and knowledge routing, with an optional minimal repository-specific skill portfolio.

Search the web with LLM-optimized results via the Tavily CLI, returning relevant snippets, relevance scores, and metadata. Supports domain filtering, time ranges, and multiple search depths.

Conduct deep, multi-source web research with citations using the Tavily CLI, producing structured reports for comparisons, market analysis, and literature reviews.

Summarizes, condenses, or extracts text, transcripts, and key points from URLs, articles, web pages, PDFs, podcasts, YouTube videos, and local files. Acts as a fallback transcription tool when the user asks to transcribe, digest, recap, or get the gist of a link, video, or document.

Turns strategy-first Madden NFL YouTube videos into structured, game-ready notes covering meta snapshot, formations, plays, adjustments, and provenance, using transcripts and a curated source list.

Transforms chaotic notes, voice dumps, and meeting transcripts into a structured, interlinked knowledge base inside Obsidian by extracting entities, generating wikilinks, and surfacing forgotten context.

Generates a personalised nutrition report from consumer genetic data (23andMe, AncestryDNA, VCF) by interrogating nutritionally-relevant SNPs and producing actionable dietary guidance, all computed locally.

Performs HLA allele genotyping from WGS/WES VCF data, producing a structured markdown report and machine-readable JSON results.

Audits a paper draft by tracing every numeric claim and assertion back to result files on disk, flagging untraced, mismatched, stale, uncited, and phantom references.

Analyzes evaluation results by requiring a stated hypothesis before examining data, then compares expectations to actual result files to prevent post-hoc rationalization.

Explains chemical principles, reactions, and safety, and supports analysis, handling, and storage guidance.

Provides aerospace domain knowledge, terminology, and guidance on aircraft and spacecraft principles, regulations, and mission planning.

Runs a full website SEO audit by crawling up to 500 pages, detecting business type, delegating to specialist subagents, and producing a prioritized action plan with a health score.

Searches the web for information, current content, and news when users need up-to-date results.

Conduct multi-pass web and academic research, cross-check claims, and produce a cited report saved to file and memory.

Routes research requests to appropriate modes and runs multi-agent deep research with evidence chains, plus optional monitoring.

Searches the web for text, news, images, videos, and books using the duckse CLI, with options for time filters, regions, safe search, and JSON output.

Analyzes competitive landscapes using Porter's 5 Forces, positioning matrices, and battlecard generation to secure market position.

Routes quality reviews to specialized critic agents based on file type or flags, covering peer review, code review, and manuscript polish.

Computes bibliometric indices for AI/NLP conferences, including QQE, average/median citations, and citation inequality, from annual publication and citation data.

Evaluates the throughput and resource allocation efficiency of RIS-aided mobile edge computing systems by measuring the total computation task bits successfully completed under varying network conditions.

Benchmarks medical AI agents on synthetic EHR tasks, measuring success rates for data retrieval, temporal aggregation, and threshold-based conditional logic with exact-match scoring.

Benchmarks energy-function-based safe control algorithms on the BIS (Benchmark of Interactive Safety) dataset, scoring safety, efficiency, and hybrid performance in human-robot and robot co-working scenarios.

Strategic decision retrospective — review research decisions against actual outcomes, update confidence levels

Amend approved research artifacts through a bounded alignment-gated workflow

AI legal document generation powered by CellCog. Contracts, NDAs, terms of service, privacy policies, freelance agreements, employment contracts, compliance reviews, legal research. Precision legal drafting from natural language prompts. Not legal advice.

Provides PhD-level mathematical analysis for machine learning, including optimization theory, statistical learning theory, and convergence proofs.

Get research-backed architecture advice for your knowledge system. Describe your use case, constraints, and goals — get specific recommendations grounded in TFT research with rationale for each decision. Triggers on "/recommend", "what would you recommend", "architecture advice", "knowledge system for".

Use to confirm clinical ethics and research-integrity requirements for an NEJM submission — IRB/ethics-committee approval and informed consent per the Declaration of Helsinki, ICMJE conflict-of-interest disclosures and authorship criteria, the role-of-the-funding-source statement, and the ICMJE data-sharing statement.

Use when auditing an ASE (IEEE/ACM Automated Software Engineering) research-track submission for HotCRP readiness, covering the ACM acmart sigconf template and the 10+2 page budget, double-anonymous review, the mandatory Data Availability Statement, the early-rejection stage before rebuttal, and desk-reject triage before the deadline.

Use when auditing an ESEC/FSE research-track submission for HotCRP readiness, covering the mandatory paper-registration step, the ACM Primary Article Template page budget, double-anonymous "heavy" anonymity, the open-science Data Availability statement, PACMSE publication, and desk-reject triage before the AoE cutoff.

Routes a research topic through a structured two-phase workflow: generate an extensible outline, then fan out parallel web-search agents to investigate each item into validated JSON, producing a complete markdown report with table of contents.

Summarize or extract text/transcripts from URLs, podcasts, and local files (great fallback for “transcribe this YouTube/video”).

Jira

Use quando o usuário solicitar a ativação do Gemini CLI para revisão de código, revisão de plano ou processamento de contexto amplo (>200k). Ideal para análise abrangente exigindo janelas de contexto grandes. Usa Gemini 3 Pro por padrão para raciocínio e codificação de última geração.

Run a structured, multi-source deep-research loop on a question until the answer stops changing.