Skills for your AI
Generates structured, numbered Architectural Decision Records in docs/adr/ with context, decision drivers, alternatives, and consequences, grounded in repository state. Use when a team decision needs documenting, a proposed option needs a review document, or e
Refactors bloated agent instruction files into a minimal root file with linked, categorized subfiles using progressive disclosure. Use when an AGENTS.md, project instructions file, or similar agent instruction file has grown too long and needs reorganizing.
Supports research scientists across the algorithm development lifecycle, from data preprocessing and feature engineering through model selection, implementation, evaluation, optimization, validation, and documentation. Use when the user needs help cleaning dat
Analyzes algorithm time and space complexity, suggests efficiency improvements, data structure and memory optimizations, parallelization plans, redundancy elimination, heuristic and genetic algorithm design, profiling and parameter tuning, trade-off analysis,
Designs scalable Angular 15+ architectures, optimizes RxJS and NgRx state management, plans micro-frontends with Module Federation, and builds migration, performance, testing, and Nx monorepo strategies. Use when planning Angular architecture, fixing memory le
Produces cloud-native architecture diagrams and documentation in Mermaid syntax, covering system context, components, deployment, data flow, sequences, NFR analysis and phased plans. Use when the user asks for architecture diagrams, an architecture document, d
Produces architecture modernization strategy documents covering service decomposition, event-driven design, API gateways, data migration, observability, and performance. Use when a user needs to break up a monolith, model business processes as events, draft Op
Drafts and reviews automated testing scripts, data, environments, CI/CD execution plans, coverage assessments, refactoring plans, and reports for developers. Use when writing or improving test scripts, generating test data, setting up test environments, automa
Extracts resources and JavaScript from any installed Electron app (`.asar` bundle), restoring original sources from `.js.map` files when available or formatting minified code with Prettier otherwise. Use when user wants to "extract Electron app", "decompile El
Automates Benchling lab data management via its Python SDK and REST API for registry, inventory, ELN, workflow, and event operations. Use when creating or updating Benchling entities, containers, notebook entries, workflow tasks, or event-driven integrations.
Turns a project idea into a complete, production-ready system plan — architecture, components, data flow, file-level execution and dependencies — as an interlinked markdown blueprint wiki. Use when a user brings a new project idea, asks to design an architectu
Executes structured coding workflows (Loop, Debug, Express, Main) with strict correctness, verified facts, and minimal tool use. Use when the user reports a bug, requests a refactor, adds a feature, or asks for a small local code change.
Generates C4 model architecture diagrams as Mermaid markdown from user descriptions or codebase exploration, at Context, Container, Component, Deployment, or Dynamic level. Use when the user asks for an architecture diagram, wants a codebase analyzed into syst
Designs, builds, and tests chatbots covering architecture, NLP intent recognition, dialog management, response generation, error handling, API integration, authentication, multilingual support, and testing. Use when a developer needs chatbot architecture, inte
Mines offline Telegram Desktop chat exports (result.json) for unmet market needs and product opportunities using chunked streaming and verbatim quote grounding. Trigger phrases: mine chat export, telegram result.json, find unmet needs, analyze telegram chat.
Retrieves bioactive molecule, target, bioactivity, structure, and drug data from ChEMBL via its Python client. Use when the user asks for ChEMBL lookups, molecule or target searches, IC50/Ki/EC50 data, similarity or substructure searches, drug mechanisms and i
Develops and verifies Clojure solutions in the REPL before touching any files, covering root cause fixes, refactoring, architecture review, and failing-test debugging. Use when fixing Clojure bugs, refactoring functions, reviewing code for best practices, or d
Builds and operates CocoIndex Python flows for AI data pipelines, covering requirements gathering, dependency and environment setup, flow code generation, custom functions, running flows, and debugging. Use when a developer wants to index files or data into a
Analyzes an existing codebase's patterns and conventions and produces a complete implementation blueprint for a requested feature. Use when the user asks to design a new feature, plan an implementation, map data flow, or produce an architecture blueprint for a
Reviews code, troubleshoots errors, inspects stack traces, writes test cases, guides environment setup, version control, tooling, performance profiling, and error handling for developers. Use when a developer shares code, an error message, a stack trace, a fai
Traces a codebase feature from entry points through all layers to data storage and produces a structured analysis report. Use when asked to analyze a feature, find entry points, trace a call chain, map architecture, or document implementation details.
Analyzes provided code to propose refactorings for clarity, performance, and maintainability, covering method extraction, duplication, conditionals, error handling, data structures, dead code, design patterns, and parameter objects. Use when a developer shares
Reviews pull requests and code diffs for security, performance, design, and test coverage following Sentry engineering practices, producing draft review comments. Use when given a PR URL or diff to review, when checking Python/Django or TypeScript/React patter
Reviews code for style, readability, performance, security, architecture, error handling, dependencies, tests, version control, and quality metrics, returning findings and fixes. Use when the user asks for a code review, style check, security scan, performance
