Claude Research

Read, create, combine, split, rotate, OCR, watermark, secure, or extract content from PDF files. Use when a PDF is a primary input or requested output. Not for LaTeX source editing or Word and spreadsheet deliverables.

Create, read, edit, or convert Microsoft Word documents while preserving professional document structure. Use when a .docx file is a primary input or requested output, including tracked changes, comments, images, headings, page furniture, or polished reports and letters. Not for PDFs, spreadsheets, Google Docs, or unrelated coding tasks.

Diagnose and fix residual TikZ label, arrow, box, and Bézier-path collisions with geometric calculations. Use when a generated .tex figure has overlapping labels or crossed arrows. Not for generating a new figure; use $figure or the upstream deck workflow.

Create, read, edit, clean, format, chart, or convert spreadsheet files while preserving spreadsheet-native deliverables. Use when .xlsx, .xlsm, .csv, or .tsv data is the primary input or a spreadsheet is the requested output. Not for Word documents, HTML reports, standalone scripts, database pipelines, or Google Sheets API work.

Capture and integrate improvement ideas for the shared Claude Code and Codex infrastructure. Use when recording an infrastructure enhancement for later review rather than implementing it immediately.

Compile one specified LaTeX document, resolve build errors, audit citations, and report build quality. Use when a .tex source must be built or a concrete compilation failure repaired. Not for corpus-wide build checks, visual polish after a clean build, or creating a project; use $latex-health-check, $latex-polish, or $latex-scaffold.

Propose and, after approval, execute content-aware file reorganization or deduplication using a local Mac Mini model with a reversible undo record. Use when sorting a folder by file contents or finding exact and near duplicates. Always dry-runs first.

Create, accept, or update the persistent `.context/ai-handoff.md` state shared by AI sessions and machines. Use when continuing in a new Claude or Codex session, transferring work between clients, or recording ownership and restart state.

Run an interactive one-question-at-a-time oral drill for research defence or active-recall study, escalating around weak answers and ending with a study sheet. Use when the user asks to be grilled, quizzed, or prepared for a viva, job talk, seminar, or exam. Not for a written critique; use $devils-advocate or a review agent.

Extract citations from a PDF and generate a validated .bib file. Use when the user asks to extract citations from a PDF and generate a validated .bib file. Reads the PDF, identifies referenced works, constructs BibTeX entries, and verifies metadata.

Proofread a LaTeX academic paper across eleven language, notation, citation, and consistency categories. Use when the requested task is bounded proofreading rather than substantive rewriting, clarity stress-testing, or venue compliance.

Download, split, and deeply read an academic PDF that is not available through Paperpile. Use when a long external PDF needs page-wise ingestion. For Paperpile items, use the Paperpile text-extraction route instead.

Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata. Use when running the scheduled knowledge-corpus promotion or previewing candidates interactively. Not for consolidating overlapping concepts; use $wiki-merge.

Filter a LaTeX project's bibliography to the entries actually cited, with Overleaf-source resolution, temporary staging, validation, approval-gated canonical replacement, and compilation. Use before submission or after large revisions.

Save a compact continuation checkpoint so current work survives context compaction or a later session. Use when the active context is becoming long or work must pause without changing client or machine ownership. For an explicit cross-session/client transfer, use $handoff.

Compare two LaTeX files, project directories, or Git revisions and produce human-readable plus machine-readable severity-graded semantic changes. Use when determining what changed between manuscript versions or preparing a revision audit bundle. Not for proofreading one version; use $proofread.

Formalize a self-authored lemma or theorem in Lean 4/mathlib and require a clean `lake build` without `sorry`. Use when the mathematical claim can be stated faithfully and machine-checked. For numerical falsification or symbolic algebra, use $numerical-check or $symbolic-check.

Write clear, detailed mathematical proofs for academic papers. Use when the user asks to prove a result, derive an equation, justify a claim analytically, or expand a proof sketch into a full proof. Also trigger on "prove", "show analytically", "derive", "justify mathematically", or "write a proof".

Deliver a structured post-mortem after incidents, mistakes, or stuck sessions. Use when the user requests a structured post-mortem after incidents, mistakes, or stuck sessions.

Create and maintain Python environments and dependencies with uv. Use when installing packages, creating a virtual environment, resolving Python dependency state, or migrating away from pip. Not for general Python coding.

Fork an existing paper into a concurrent second-venue submission variant with policy checks, separate Overleaf ownership, format conversion, budget refit, QA, and Atlas/Vault writeback. Use when submitting the same work concurrently to another permitted venue. Not for preprints, ordinary retargeting, or camera-ready work.

Merge an overlap cluster of Vault concepts into one canonical concept, rewrite wikilinks, and preserve a dry-run/apply boundary. Use when $wiki-curate has identified a winner and fold-in slugs. Not for general wiki auditing or automatic article promotion.

Create an academic presentation as a LaTeX Beamer source and reviewed PDF with an original theme. Use when the requested deliverable is a conference, seminar, or lecture deck in Beamer. Not for PowerPoint or RevealJS; use $pptx or $quarto-deck.

Guide for creating MCP servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

Generate a Reveal.js HTML presentation from Markdown. Use when the user asks to generate a Reveal.js HTML presentation from Markdown.

Create a timestamped research-session progress log that records completed work, evidence, decisions, and next steps. Use when closing or checkpointing substantive project work that needs a durable chronology. Not for saving one reusable finding; use $store-insight.

Verify a self-authored mathematical result end to end by routing claims across adversarial review, numerical falsification, symbolic or CAS checks, and Lean, then aggregating one report. Use when a theorem, proposition, conjecture, or paper-wide mathematical argument needs the appropriate combination of verification methods. For one known method, use $numerical-check, $symbolic-check, or $lean-check directly.

Audit the Research Vault concept wiki read-only for fragmentation, missing tags, write-only concepts, and draft or anatomy conformance, producing a Markdown report. Use when assessing wiki quality before maintenance. Not for promoting or merging concepts; use $wiki-grow or $wiki-merge.

Compare a project .bib against a Paperpile project/topic folder to find uncited papers or unfiled entries. Use when the user asks to compare a project .bib against a Paperpile project/topic folder to find uncited papers or unfiled entries.

Convert an accepted anonymous-submission LaTeX paper (AAAI/AIES/ACM-style) to camera-ready and implement the accepted reviews. Use when a paper is accepted with no rebuttal and you need to de-anonymize, add copyright, turn on section numbering, implement each reviewer's minor revisions, optionally move proofs to a non-counted appendix, and QA. Not for R&R/revise-and-resubmit (use strategic-revision --external) or for preprints (use preprint).

Bootstrap a new research project. Interview for details, scaffold directory structure, create Overleaf symlink, initialise git, and create project context files.

Conduct an adaptive structured interview that elicits tacit knowledge, requirements, preferences, or decisions and summarizes them explicitly. Use when the needed information is in the user's head and cannot be recovered from project files. Not for adversarial oral examination; use $grill-me.

Inspect a cleanly compiling LaTeX document for source pathologies and rendered visual defects by linting and viewing selected PDF pages. Use when compilation succeeds but title pages, floats, tables, figures, or layout still need publication-quality review. Not for basic compilation health; use $latex-health-check.

Extract and compare data-collection methods across a set of empirical papers. Use when the user needs a cross-paper methods matrix or wants to assess how a literature gathers evidence.

Create a concise presentation deck that communicates a project's current state, decisions, evidence, risks, and next steps. Use when preparing a project update for collaborators, supervisors, or stakeholders. Not for a research talk centered on one paper; use $talk-deck.

Update current-focus.md with a structured session summary. Use when the user asks to update current-focus.md with a structured session summary.

Edit draft or auto-generated prose to match an existing personal voice profile. Use when transforming generic text into the profiled author's wording, cadence, and style. Not for creating the profile or matching a journal; use $voice-analyzer or $journal-voice.

Design or audit the identification strategy for an observational study. Use when the task concerns estimands, causal assumptions, threats to identification, or defensible research design rather than model implementation.

Deliver an end-to-end analysis pipeline: EDA, estimation, or publication output. Use when the user requests an end-to-end analysis pipeline: EDA, estimation, or publication output.

Archive an exported Claude Code insights HTML report and turn it into a timestamped Beamer presentation usable from either client. Use when the insights export is the source artifact for a presentation. Not for a general research talk; use $talk-deck.

Create a large-format academic poster in LaTeX using beamerposter, tikzposter, or baposter. Use when the requested deliverable is a conference or research poster rather than sequential presentation slides. For slide decks, use $beamer-deck.

List recent meetings and voice memos. Use when the user asks "what meetings did I have", "show my recent recordings", "any meetings today", "list my voice memos", or wants an overview of their meeting history. Also use when they need to find a specific meeting by browsing rather than searching.

Interactive meeting preparation — builds a relationship brief and talking points before a call. Use when the user says "prep me for my call with", "I'm meeting with X", "prepare me for", "what should I bring up with", "meeting prep", "get ready for my call", or wants to review history with someone before a meeting.

Create, revise, and evaluate reusable AI workflow skills, including trigger-quality tests. Use when authoring a new skill, repairing an existing skill, or measuring whether its metadata routes correctly.

Extract reusable knowledge from the current session into a persistent skill. Use when you discover something non-obvious, create a workaround, or develop a multi-step workflow that future sessions would benefit from.

Lint an existing project knowledge base for contradictions, uncited claims, missing connections, stale articles, and orphaned concepts. Use when checking the integrity of already compiled knowledge before relying on or expanding it. Not for compiling raw sources; use $compile-knowledge.

Convert an existing Markdown manuscript draft into a buildable LaTeX project with canonical structure and build configuration. Use when prose already exists but no LaTeX source tree does. Not for checking template drift in an existing project; use $latex-template.

Compare a project's LaTeX preamble and conventions against the canonical working-paper template. Use when diagnosing template drift without converting venue formats or editing prose.

Generate a daily digest of today's meetings and voice memos — key decisions, action items, and themes across all recordings. Use when the user asks "recap my day", "what happened in my meetings today", "daily summary", "what did I discuss today", "any action items from today", or wants a consolidated view of the day's conversations.

Automate browser interactions and inspect or test web pages through Playwright commands. Use when the task requires deterministic browser navigation, screenshots, selectors, or Playwright-test work. For a full local-webapp test workflow, use $webapp-testing.

Install or reconcile research-project safety rules and protected folder boundaries for data, code, outputs, and paper sources. Use when a project lacks guardrails or is being migrated into the governed structure. Not for auditing the full research scaffold; use $audit-project-research.

Deliver a mid-draft adversarial review of a paper — runs paper-critic + domain-reviewer + claim-verify + blindspot in parallel, optionally adds clarity-reviewer, then auto-synthesises into a prioritised revision plan. Use when the user requests a mid-draft adversarial review of a paper — runs paper-critic + domain-reviewer + claim-verify + blindspot in parallel, optionally adds clarity-reviewer, then auto-synthesises into a prioritised revision plan. Distinct from pre-submission-report (final-gate kitchen sink, 14 checks) — this is the active-drafting feedback loop. Triggers: 'review my draft', 'adversarial review', 'cluster review', 'mid-draft critique', 'feedback before pre-submission'.

Check whether the current session has coherent guidance, focus, handoff state, working-tree ownership, and enough context to proceed safely. Use when starting, resuming, or diagnosing a confused session. Not for auditing the machine or deployed AI infrastructure; use $system-audit.

Use SymPy to prove or refute a self-authored algebraic identity, derivative, limit, comparative-static sign, or closed form. Use when exact symbolic manipulation can settle the claim. For parameter sweeps or full theorem proving, use $numerical-check or $lean-check.

Generate structurally realistic synthetic datasets for pipeline prototyping, test coverage, or prospective power analysis. Use when code or design must be exercised before real data are available or accessible. Never substitute synthetic records for governed raw data.

Analyze representative writing samples into a portable personal voice profile and style guide. Use when establishing voice-matched editing for a new project or refreshing an outdated profile. Not for a target journal's editorial style; use $journal-voice.

Exercise and verify a local web application through Playwright, including user flows and rendered behaviour. Use when testing a running local app rather than issuing an isolated browser command. For ad hoc automation, use $playwright-cli.

Adversarially challenge research assumptions, mechanisms, and arguments in writing. Use when stress-testing a claim or design before committing to it. Not for an interactive oral drill or a full referee report; use $grill-me or a review agent.

Search meeting transcripts and voice memos for people, topics, decisions, commitments, or remembered ideas. Use when the user asks what was discussed, decided, or said across their meeting history. Not for preparing an upcoming meeting; use $meetings-prep.

Weekly meeting synthesis — themes, decision arcs, stale commitments, and what deserves your attention next week. Use when the user says "weekly review", "what happened this week", "weekly summary", "recap my week", "any outstanding items", "week in review", or at the end of a work week.