handbook
handbook には garg-aayush から収集した 19 個の skills があり、リポジトリ単位の職業カバレッジとサイト内 skill 詳細ページを表示します。
このリポジトリの skills
Generate commit messages for staged git changes. Use when the user asks to generate a commit message or wants help writing commit messages.
Debug errors using logs and stack traces. Use when the user shares an error message, crash log, or runtime issue to fix.
Explains code with visual diagrams and analogies. Use when explaining how code works, teaching about a codebase, or when the user asks "how does this work?"
Explain errors without editing files. Use when the user wants to understand an error or crash but not immediately fix it.
Analyze code for logical errors, runtime bugs, and edge cases. Use when the user wants a code review, asks to find bugs, or wants to check code correctness.
Image manipulation via a bundled script, resizing, rotation, flipping, cropping, padding, format conversion (JPEG/PNG/WebP/TIFF/HEIC), transparency operations (remove/replace/extract/blend), grayscale conversion, auto-cropping borders, and file size optimization.
Convert documents to Markdown. PDFs go through the LlamaParse API (best quality, costs credits, tier selectable) or a local PyMuPDF engine (free, offline, no API key); other formats (DOCX, PPTX, XLSX, HTML, EPUB) are converted locally with markitdown.
Convert a daily log file's bullet points into a structured checklist (checkbox, short title, description, status per item), editing the file in place.
Remove AI-generated code slop (excessive comments, over-defensive code, type workarounds, style mismatches) from specific files or all changes on a branch. Conservative, preserves functionality and matches existing codebase style.
Reformat raw Fireflies meeting notes into structured markdown with Key Takeaways, timestamped Notes sections, and per-person Action Item checklists, overwriting each file in place. Light copy-edits only, no summarizing or rewriting.
Generate and edit images with Google's Gemini image models (Nano Banana family) via a bundled script. Text-to-image, image editing, and multi-reference composition at 1K/2K/4K with aspect-ratio control across lite/flash/pro model tiers. Do NOT read image files and do NOT write new image-generation code; always call the bundled script, passing image paths via --input-image.
Stress-test an existing implementation plan by challenging assumptions, probing edge cases and failure modes, and resolving each branch of the decision tree, then update the plan with what was agreed.
Interview the user to uncover intent, goals, and constraints for a new idea, then capture the result as a structured brief to carry into planning.
Generate structured notes from a transcript, Slack summary, huddle recording, or call log. Handles messy auto-generated transcripts, reconstructs meaning from context, and writes the finished notes to a markdown file.
Create or improve a schematic diagram (architecture, pipeline, flow, sequence, ER, state) as a mermaid block in markdown, with dark-first theming, role colors, anti-clipping labels, and render validation. Always prefer mermaid over ASCII art.
Generate an onboarding cheat-sheet for an unfamiliar repository, covering what it is, the tech stack, architecture, local dev, deployment / CI/CD, databases, and critical engineering points. Produces a single grounded markdown file the user keeps and updates.
Review Python code for bugs, edge cases, and pitfalls specific to ML, data, API, and deployment contexts (data leakage, tensor shapes, device placement, missing timeouts, hardcoded secrets, resource leaks). Reports findings in chat without editing files.
Restructures a raw, unformatted prompt into a clean, XML-tagged prompt without altering its content. Outputs the structured prompt only; does not execute the task the prompt describes.
Fetch a Jira ticket using the Atlassian MCP plugin and extract key details into a structured summary. Use when the user wants to retrieve, view, or get information about a Jira ticket.