Skills for your AI
Adapts existing PyTorch, TensorFlow, and HuggingFace training scripts to run distributed on Ray Train across multiple GPUs or nodes, including hyperparameter tuning, checkpointing, and cluster status checks. Use when the user wants to scale training to multi-G
Stores a household dietary profile and vets DoorDash carts against it before checkout, flagging allergens, diets and dislikes. Use when setting up or updating a dietary profile, after any cart add/remove/reorder, when requesting a checkout URL, or when orderin
Coordinates group DoorDash orders by building a merged cart, attributing items to eaters, splitting costs with fee proration, and tracking payer rotation. Use when the user asks to order food for a team, parse a pasted chat thread into an order, edit items by
Answers questions about DoorDash spending and ordering history by querying the append-only audit log at ~/dd-guard/audit.jsonl. Use when the user asks what DoorDash orders were placed, what they cost, what the agent did in past sessions, what log event types m
Saves and recalls DoorDash orders as named playbooks with cart diff and drift detection before checkout. Use when the user names a saved order, wants to save a recent order, list or remove playbooks, handle a stale playbook, or run first-time setup.
Builds, trains, tests, and maintains customer service chatbots for e-commerce stores, covering training content, NLP flows, integrations, testing, analytics, personalization, order flows, support automation, feedback, loyalty, multi-language, and maintenance.
Analyzes eLearning user feedback into categorized insights, sentiment, bugs, usability and performance findings, comparisons, recommendations and stakeholder reports. Use when given course or platform feedback, behavioral data, or requests to triage issues, co
Guides support representatives through the full escalation lifecycle—triage, documentation, internal and customer communication, follow-up, closure, root cause analysis, templates, and metrics. Use when a rep describes a customer issue needing escalation, draf
Helps support specialists identify, assess, communicate, document, and improve customer escalations by drafting communications, matrices, procedures, training, and trend reports. Use when a conversation shows frustration or repeated contact, when an escalation
Helps call center supervisors define escalation criteria and levels, build escalation procedures and timeframes, document cases, analyze trends, sentiment and performance, and create agent training. Use when a supervisor needs escalation guidelines, triage cri
Plans and builds production-ready FastAPI endpoints with async SQLAlchemy, Pydantic v2, dependency-injected auth, and pytest tests. Use when adding or extending an endpoint in an existing FastAPI project, or when asked to plan, implement, or test a FastAPI res
/cs:fda-qsr-audit-prep <scope> — FDA 21 CFR 820 (QSR / QMSR) audit 6-question forcing interrogation. Post-Feb 2026 substantially harmonized with ISO 13485. Use before annual internal QSR audit, pre-FDA-inspection readiness, or Form 483 response.
Turns raw customer feedback into decision-ready insights—sentiment, topics, categories, trends, satisfaction drivers, competitor benchmarks, and root causes. Use when a manager provides reviews, surveys, or support tickets and asks for sentiment breakdowns, th
Turns scattered customer feedback into structured, prioritized insights — sentiment, topics, categories, summaries, trends, priorities, competitor comparisons, feature requests, and visualizations. Use when analyzing customer reviews, support tickets, surveys,
Converts user-provided files between formats via the ChangeThisFile service, lists supported conversions, and handles conversion errors. Use when the user supplies a file or URL and a target format, asks whether a conversion is possible, or a conversion attemp
Fetches AI-powered match predictions for Premier League and Champions League matches, including scores, next goal scorer and corner counts. Use when the user asks for matchweek predictions, a specific matchweek, predictions filtered by home or away team, or as
Use when the user asks to capture a full-page screenshot, long screenshot, or complete page capture of a web page. Handles SPA scroll containers, lazy-loaded images, and very tall pages via Chrome DevTools Protocol with zero external dependencies.
Fetches the open GitHub PR for the current branch, lists its review comment threads, applies fixes for the comments the user selects, and tracks which ones are already addressed. Use when the user asks to see, triage, or fix review comments on their PR.
Sets up GitHub as a chat channel by copying and registering an adapter, installing a pinned package, collecting credentials, and wiring the channel to an agent. Use when the user wants the agent to participate in GitHub pull request or issue comment threads.
Docs-anchored grilling session — challenges a plan against the project's existing language (CONTEXT.md) and recorded decisions (docs/adr/), and updates those files inline as terminology and decisions crystallise. Use when user wants to stress-test a plan again
Guides hackathon teams through context gathering, concept ranking, time-boxed scheduling, mid-event re-scoping, and pitch preparation. Use when a team needs a hackathon concept selected, a phase schedule, triage after falling behind, or a demo and pitch plan.
/em:hard-call — Framework for decisions with no good options. Use when every option is painful and a structured 10/10/10 + regret-minimization pass is needed — e.g. choosing between a layoff and a down round, or killing a beloved product line.
Provides step-by-step guidance for hardware and software configuration, installation, troubleshooting, and optimization tasks. Use when a help desk technician needs instructions for installing components, configuring BIOS/UEFI, installing software or drivers,
Use when Codex should act as the Healthcare Customer Service specialist from Agency Agents. Empathetic healthcare customer service specialist for patient support, billing inquiries, appointment management, insurance questions, complaint resolution, and seamles
