HezaoHezao/poirot

Produces structured, peer-review-quality analyses of academic papers, covering methodology assessment, strengths, weaknesses, and actionable recommendations.

Enables searching and retrieving academic papers from arXiv using a free REST API. Ideal for researchers needing quick access to papers by keyword, author, or category.

Generates dark-themed SVG architecture, cloud, or infrastructure diagrams as standalone HTML files. Ideal for system architectures, microservices, and deployment visualizations without dependencies.

Monitor blogs and RSS/Atom feeds using blogwatcher-cli. Supports automatic feed discovery, OPML import/export, and read/unread article tracking for daily updates.

Generates a detailed user profile through a structured 5-8 round onboarding conversation, adapting to the user's language and style. Use when setting up a new agent or refreshing an agent's understanding of the user.

Transforms data into visual charts by intelligently selecting chart types, extracting parameters, and generating chart images using Python matplotlib/plotly.

Generate comprehensive documentation for software projects, including READMEs, API references, architecture diagrams, and developer guides. Use when documenting codebases, improving existing docs, or creating changelogs.

Perform code reviews on local changes before pushing, or review open pull requests on GitHub. Automates diff analysis, inline commenting, and formal reviews via gh or REST API.

Codebase Inspection analyzes a repository for lines of code, language breakdown, file counts, and code-vs-comment ratios using pygount. Use it when you need a quick read on codebase size or composition.

Generate minimal SVG diagrams as self-contained HTML files with automatic light/dark mode. Ideal for educational visuals, process flows, and technical illustrations.

Generates structured analysis frameworks and synthesizes data into professional, consulting-grade research reports covering market analysis, consumer insights, financial analysis, and more, following McKinsey/BCG standards.

Analyzes user-provided Excel (.xlsx/.xls) or CSV files using DuckDB — an in-process analytical SQL engine. Supports schema inspection, SQL querying, statistical summaries, and result export.

A systematic research methodology that guides you through broad exploration, deep dives, diversity validation, and synthesis checks to ensure thorough web research before content creation. Use this skill whenever a single search is insufficient and you need multiple perspectives, data, and expert insights.

Helps agents find relevant skills from Poirot's builtin library by keyword search. Use when starting a specialized task to check for existing skill workflows.

Creates distinctive, production-grade frontend interfaces with unique aesthetics, avoiding generic AI-generated designs. Use for crafting memorable, visually striking UIs with meticulous attention to detail.

GitHub Authentication Setup configures HTTPS tokens, SSH keys, or gh CLI auth so an agent can work with GitHub repositories, pull requests, issues, and CI. Use it before cloning, pushing, or calling the GitHub API.

Performs multi-round deep research on GitHub repositories using API calls, web search, and page browsing to generate comprehensive markdown reports.

GitHub Issues Management creates, searches, labels, assigns, comments on, and closes GitHub issues using gh or the REST API. Use it for issue triage and basic issue administration.

GitHub Pull Request Workflow covers branching, committing, opening PRs, checking CI, reviewing, merging, and cleaning up branches. Use it when managing a GitHub PR from first commit to merge.

GitHub Repository Management creates, clones, forks, and configures GitHub repositories. Use it to manage remotes, releases, settings, secrets, and deletion from the command line or API.

Newsletter Generation researches and writes professional newsletters, digests, and roundups from multiple sources. Use it when you need a ready-to-publish Markdown briefing with summaries, commentary, and links.

Node Inspect Debugger drives Node.js's built-in inspector to set breakpoints, step through code, and evaluate expressions. Use it when logs are not enough or you need to inspect a running process, call stack, or profile.

OSINT Investigation cross-references public records from SEC, sanctions, courts, property, archives, and news. Use it to resolve entities, build evidence chains, and assess confidence in findings.

Generates professional PowerPoint presentations from structured slide plans and content. Supports multiple visual styles including dark-premium, gradient-modern, and minimal-swiss.

Generates detailed markdown implementation plans with step-by-step tasks, code examples, and testing commands for AI coding assistants or developers.

Automated pre-commit verification pipeline that runs security scans, quality gates, and a fresh-eyes review with auto-fix capability. Use it to catch issues before your code lands in a shared branch.

Python Debugger uses pdb and debugpy to pause Python code, inspect state, and attach to running processes. Use it for failing tests, post-mortem inspection, or long-running jobs that need remote debugging.

Research Paper Writing Pipeline guides ML/AI paper work from experiment design through analysis, drafting, review, and submission. Use it when preparing a venue submission or iterating on results and revisions.

Reviews recent code changes with three sequential lenses (reuse, quality, efficiency) to identify and apply improvements, reducing technical debt without overwhelming the developer.

Guides the authoring of SKILL.md files for Poirot skills, covering required frontmatter, description standards, size limits, writing principles, and peer-matched structure. Use when creating or editing builtin or user-local skills.

Guides the creation of new agent skills by walking through an iterative loop of drafting, testing, evaluating, and refining. Use it when building or improving skills for the Poirot agent platform.

Guides AI agents in cross-referencing sources to verify factual claims, ensuring information credibility through multi-source verification and source credibility annotation.

Run disposable experiments to validate technical feasibility, compare approaches, and surface unknowns before committing to a full build. Spikes are thrown away after providing insight.

Subagent-Driven Development executes implementation plans one task at a time with a spec review and a code-quality review after each task. Use it for planned work where tasks are mostly independent and need test-driven implementation.

A disciplined 4-phase debugging methodology that forces root cause investigation before attempting any fix. Use for test failures, production bugs, or unexpected behavior to avoid random patches.

Systematic Literature Review searches arXiv papers on a topic, extracts structured metadata, and synthesizes themes, convergences, disagreements, and gaps. Use it for literature surveys, annotated bibliographies, and topic overviews across multiple papers.

A rigorous test-driven development workflow enforcing red-green-refactor cycles: write failing tests first, watch them fail, then write minimal code, and refactor. Ideal for AI coding agents to ensure code quality and prevent untested production code.