Results for “pre-pr”

Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.

Implement API abuse detection using token bucket, sliding window, and adaptive rate limiting algorithms to prevent DDoS, brute force, and credential stuffing attacks.

Configure and deploy Palo Alto Networks next-generation firewalls with App-ID, User-ID, zone-based policies, SSL decryption, and threat prevention profiles for enterprise network security.

Explains MariaDB Connector/Node.js specifics: Promise vs Callback API, placeholders, BigInt handling, batch inserts, query streams, pooling, transactions, and prepared statements, for writing or reviewing Node.js code that uses the `mariadb` package.

Complete meeting lifecycle skill — preparation before and follow-up after any professional or personal meeting. Use this skill whenever the user mentions an upcoming meeting, wants to prepare for a meeting, needs an agenda, wants to brief themselves before talking to someone, needs to write up meeting notes, wants to send a follow-up email after a meeting, needs to capture action items, or says anything like "I have a meeting with...", "I just got out of a meeting", "help me prepare for...", "I need to follow up on...", or "what should I ask in this meeting." Also trigger for specific meeting types: job interviews, performance reviews, 1:1s, sales or client calls, and difficult conversations. Both modes are covered: PRE-meeting preparation and POST-meeting follow-up. Trigger proactively even when the user doesn't use the word "meeting" — if they describe a situation where they're about to talk to someone important, or just finished such a conversation and need to capture it, this skill applies.

Use this skill to prevent destructive operations when working on production systems or running agents autonomously.

Scores real estate leads based on engagement, budget, timeline, and property preference matching

Forecast any univariate time series (sales, sensors, energy, vitals, weather) zero-shot using Google's TimesFM foundation model, with point forecasts and prediction intervals from CSV, DataFrame, or array inputs.

Technical architecture guide for building a multi-purpose consent preference center. Covers per-purpose granularity, easy withdrawal under Article 7(3), version history, audit trails, and IAB Transparency and Consent Framework v2.2 integration. Includes database schema, API design, and UI component specifications.

Use when making a forward-looking claim with a checkable outcome (reply within 24h, error rate will drop, this skill will see more use) — record to state/predictions.jsonl with a review horizon so reflection can grade you later. Closes the in-the-moment double-loop.

编写评估FastText文本分类模型的Python函数，计算accuracy、F1、recall和precision指标，并处理特定格式的标签文本分割。

Audit codebase for opportunities to reduce lines of code while preserving functionality, performance, and quality

Stores verified cross-session facts, decisions, preferences, and open questions without secrets, speculation, or monitoring.

Generate PPTX presentations, DOCX documents, XLSX spreadsheets, and PDF reports from structured JSON input using PaperJSX.

Optimizes EV battery performance with charging strategies, degradation modeling, and range prediction

Designs investor pitch decks with storytelling framework, financial slides, and team presentation

Detects crop diseases from leaf images with treatment recommendations and prevention strategies

Designs ETL/ELT data pipelines with Apache Airflow, dbt, or Prefect configurations

Extended hours trading support for pre-market and after-hours sessions via Alpaca

Redige a contestacao com TODA a defesa em peca unica (CPC 335-342): preliminares do art. 337 (via preliminares-civel) + impugnacao especifica dos fatos (341 — o fato nao impugnado presume-se verdadeiro) + merito (defesa direta/indireta, fatos impeditivos/modificativos/extintivos) + provas requeridas. Prazo 15 dias uteis (335). Reconvencao na mesma peca (343 — via reconvencao). Use quando o operador disser contestacao, contestar, fui citado, defesa, vou responder a acao, prazo de contestacao, ou descrever um caso na posicao de reu.

Build sandboxed applications for secure code execution. Load when building AI code execution, code interpreters, CI/CD systems, interactive dev environments, or executing untrusted code. Covers Sandbox SDK lifecycle, commands, files, code interpreter, and preview URLs. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.

Creates a consistent minimalist hand-drawn illustrated character from user-provided photos or profile materials, then generates content visuals like knowledge cards, posters, storyboards, and presentation slides featuring that character alongside helper agents.

WordPress security hardening — nonce verification, wp_nonce_field, check_admin_referer, sanitize_text_field, sanitize_email, esc_html, esc_attr, esc_url, wp_kses, SQL injection prevention, $wpdb->prepare(), capability checks, current_user_can, authentication hardening, wp-config.php secrets, security headers, file permissions, Bedrock .env secrets, disable XML-RPC, brute force wp-login.php, CSRF protection. Invoke for: security audit, input sanitization, output escaping, nonce verification, SQL injection prevention, CSRF protection, capability checks in REST endpoints.

Build sandboxed applications for secure code execution. Load when building AI code execution, code interpreters, CI/CD systems, interactive dev environments, or executing untrusted code. Covers Sandbox SDK lifecycle, commands, files, code interpreter, and preview URLs. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.

Supplements general deposition preparation with IP-specific examination frameworks for patent, trademark, copyright, and trade secret cases. Covers witness strategies for inventors, accused infringers, licensing witnesses, and experts. Use alongside @deposition-preparation and @deposition-expert-witness when planning IP depositions, drafting outlines, or analyzing witness strategy.

Run Karpathy-style autonomous ML search on a real training repo: choose the right mode (setup, program.md, bounded loop, results interpretation, or constrained-hardware adaptation), preserve the immutable prepare.py / 300-second / val_bpb contract, and route prompt/skill eval work away to LangSmith, Promptfoo, Braintrust, or skill-autoresearch.

Use when asked for dark-mode glassmorphism cards, frosted dark hero sections, blur panels, or glass UI with gradient/glow borders via a pseudo-element mask. Prefer this over mengto-dark-glass-clean-layout when isolated glass cards/borders matter, not multi-column workspace shells; prefer over mengto-glass-dark-mode-clock when there are no clock dials; prefer over mengto-blue-laser-clean-glass-layout when gradient borders matter more than blue laser atmosphere.

Run cloud-native molecular modeling and drug-design workflows via a Python API, covering pKa prediction, docking, conformer and tautomer ensembles, molecular dynamics, and related small-molecule or protein tasks without local HPC infrastructure.

Deploy applications and websites to Vercel. Use this skill when the user requests deployment actions such as "Deploy my app", "Deploy this to production", "Create a preview deployment", "Deploy and give me the link", or "Push this live". No authentication required - returns preview URL and claimable deployment link.

Designs and governs recursive self-improvement loops where an agent mines its own failures and proposes edits to its own harness, prompts, or workflow, covering acceptance gates, diversity preservation, and the optimization ladder.

Deploy with unified build/test/deploy intent across providers using .agent-loom/deploy.yml and per-provider adapters. Load when the user asks to deploy, ship to preview, release to production, or run a provider-agnostic deploy flow. Also triggers on "deploy anywhere", "deploy to Vercel", "deploy with GitHub Actions", "preview deploy", or "ship this". Runs preflight before any deploy — stops on missing secrets. Pairs with ci-cd-and-automation for pipeline design. Ships vercel and github-actions adapters; extensible adapter interface.

Analyze AI disruption pressure across a business, map competitive exposure, and produce a 90-day defensive action plan.

Build a balanced, preference-aware meal plan for a stated wellness goal using general nutrition principles.

Finds flight deals with price prediction, flexible date search, and multi-city route optimization

Analyze AI disruption pressure across a business, map competitive exposure, and produce a 90-day defensive action plan.

Analyze AI disruption pressure across a business, map competitive exposure, and produce a 90-day defensive action plan.