Research

Perplexity AI research assistant with real-time web search, source citations, and optimized UK-centric prompts. Access live web data, current events, market research, and comprehensive analysis backed by verified sources.

Consult external AIs (Gemini 2.5 Pro, OpenAI Codex, Claude) for second opinions. Use for debugging failures, architectural decisions, security validation, or need fresh perspective with synthesis.

Deploy production recommendation systems with feature stores, caching, A/B testing. Use for personalization APIs, low latency serving, or encountering cache invalidation, experiment tracking, quality monitoring issues.

Generates, explains, and plans shell commands, bash scripts, unix pipelines, awk/sed/grep expressions. Use when user requests shell commands, bash scripts, terminal commands, unix pipes, scripting help, or command syntax assistance. Do not use for web research or document analysis.

Execute focused research tasks end-to-end with clear constraints, evidence tracking, and confidence ceilings.

Query Google NotebookLM for source-grounded, citation-backed answers from uploaded documents. Reduces hallucinations through Gemini's document-only responses. Browser automation with library management and persistent authentication.

Set up complete experimental infrastructure for all runs in a designed experiment. Orchestrates parallel generation of fine-tuning configs (via scaffold-torchtune) and evaluation configs (via scaffold-inspect). Use after design-experiment to prepare configs before running experiments.

Design and iterate on new research methods with structured checkpoints, baselines, and validation.

Research and extract an engineer's coding style, patterns, and best practices from their GitHub contributions. Creates structured knowledge base for replicating their expertise.

Use Codex CLI sandbox mode to try risky changes safely. Isolated experimentation with network disabled and directory restrictions.

Use when writing recommendation letters, reference letters, or award nominations for students, postdocs, or colleagues. Invoke when user mentions recommendation letter, reference, nomination, letter of support, endorsement, or needs help with strong advocacy, comparative statements, or evidence-based character assessment.

Quickly generate and triage research ideas with constraint clarity and confidence ceilings.

Identify and visualize research gaps, coverage, and conflicts with explicit constraints and evidence.

Systematic technology and market reconnaissance for extracting actionable intelligence from repositories, papers, and competitive landscapes.

Prepare research outputs for publication with compliance to venue rules, evidence integrity, and confidence ceilings.

[CLAUDE CODE ONLY] Leverage Codex CLI for AI peer review, second opinions on architecture and design decisions, cross-validation of implementations, security analysis, and alternative approach generation. Requires terminal access to execute Codex CLI commands. Use when making high-stakes decisions, reviewing complex architecture, or when explicitly requested for a second AI perspective. Must be explicitly invoked using skill syntax.

Use when transforming analysis/data into persuasive narratives—presenting to executives, explaining technical concepts to non-technical audiences, creating customer-facing communications, writing investor updates, announcing changes, turning research into insights, or when user mentions "write this for", "explain to", "present findings", "make this compelling", "audience is". Invoke when information needs to become a story tailored to specific stakeholders.

Use when debates are trapped in false dichotomies, polarized positions need charitable interpretation, tradeoffs are obscured by binary framing, synthesis beyond 'pick one side' is needed, or when users mention steelman arguments, thesis-antithesis-synthesis, Hegelian dialectic, third way solutions, or resolving seemingly opposed principles.

Searches the web, researches topics online, finds current information using Gemini with Google Search grounding. Use when user asks to search the web, research information, find sources, look up current events, technology updates, or gather real-time data from the internet. Do not use for local file analysis or code execution.

Use when starting a forecast to establish a statistical baseline (base rate) before analyzing specifics. Invoke when need to anchor predictions in historical reality, avoid "this time is different" bias, or establish outside view before inside view analysis. Use when user mentions base rates, reference classes, outside view, or starting a new prediction.