AI & ML

abelrguezr Bundle

How to implement LoRA (Low-Rank Adaptation) for efficient fine-tuning of large language models. Use this skill whenever the user wants to fine-tune an LLM, reduce training memory/compute requirements, implement parameter-efficient fine-tuning, or adapt a pre-trained model to a new task without retraining all parameters. Make sure to use this skill when users mention fine-tuning, LoRA, PEFT, parameter-efficient training, or want to train on limited hardware.

abelrguezr Bundle

How to fine-tune a pre-trained LLM to follow instructions and respond to tasks like a chatbot. Use this skill whenever the user wants to train an LLM on instruction-response pairs, format datasets for instruction tuning, evaluate fine-tuned model responses, or understand the complete instruction fine-tuning workflow. Make sure to use this skill when users mention instruction tuning, chatbot training, Alpaca format, Phi-3 format, or any scenario where they need to make an LLM respond to specific prompts rather than just generate text.

abelrguezr Bundle

Decode and analyze Apple device serial numbers to extract manufacturing location, date, and model information. Use this skill whenever you need to parse Apple serial numbers, investigate device origins, analyze hardware for security assessments, or understand when/where a Mac was manufactured. Trigger this for any 12-character alphanumeric Apple serial number analysis, device forensics, or hardware investigation tasks.

abelrguezr Bundle

How to identify and exploit SSH agent forwarding vulnerabilities for security auditing. Use this skill whenever you need to check for SSH agent forwarding misconfigurations, find exposed SSH agent sockets, understand agent hijacking risks, or audit SSH security. Trigger this when users mention SSH agent forwarding, SSH_AUTH_SOCK, agent hijacking, privilege escalation via SSH, or any SSH security auditing task.

abelrguezr Bundle

Android biometric authentication pentesting and bypass techniques. Use this skill whenever testing Android app security, analyzing biometric authentication implementations, or performing mobile security assessments. Trigger for fingerprint authentication testing, biometric prompt analysis, Android Keystore security reviews, or any mobile app security work involving authentication mechanisms. Don't skip this skill for Android security testing even if the user doesn't explicitly mention 'biometric' or 'fingerprint' - authentication security is always in scope.

abelrguezr Bundle

Active Directory Certificate Services (AD CS) domain escalation techniques. Use this skill whenever the user mentions AD CS, certificate templates, ESC vulnerabilities, PKI misconfigurations, certificate-based authentication attacks, or wants to enumerate/exploit certificate authority weaknesses in Active Directory environments. This covers ESC1-ESC16 techniques including misconfigured templates, enrollment agent abuse, access control issues, NTLM relay to AD CS, and certificate mapping vulnerabilities.

abelrguezr Bundle

MediaTek XFlash Carbonara DA2 hash bypass exploit for firmware analysis and security research. Use this skill whenever analyzing MediaTek device firmware, investigating Download Agent vulnerabilities, performing pre-OS security research, testing for Carbonara/heapb8 vulnerabilities, or working with mtkclient/penumbra tools. Trigger on any mention of MediaTek, XFlash, DA1, DA2, Download Agent, preloader, firmware security, or boot exploit analysis.

abelrguezr Bundle

Analyze Keras model files (.keras, .h5) for deserialization vulnerabilities, create test payloads for security research, and assess ML model security posture. Use this skill whenever the user mentions Keras models, model deserialization, ML model security, .keras files, .h5 files, TensorFlow-Keras, pickle vulnerabilities in ML, or wants to test/audit model file security. This includes CVE-2024-3660, CVE-2025-1550, gadget hunting, and Fickling-based protections.

abelrguezr Bundle

Security testing methodology for AI CLI tools (Claude Code, Gemini CLI, Warp) and MCP servers. Use this skill whenever you need to assess AI agent abuse vectors, test MCP server vulnerabilities, analyze repo-controlled configuration poisoning risks, perform authorized secrets inventory scanning, or conduct pentesting of remote MCP servers. Trigger this skill for any security assessment involving AI command-line interfaces, Model Context Protocol implementations, or LLM-powered tool execution systems.

abelrguezr Bundle

AI agent browser abuse and phishing methodology. Use this skill whenever the user mentions AI agents, browser automation, credential theft, prompt injection in browsers, agent mode phishing, hosted browser attacks, or any security testing involving AI assistants with browsing capabilities. This covers AI-in-the-middle attacks, OCR-based injections, navigation-triggered injections, trust-zone failures, and cross-site data theft via agentic browsers.

betterpromptme Skill

This prompt instructs an AI image generator to produce a hyper-realistic, high-gloss luxury editorial portrait of the woman from a reference photo while preserving her exact face and hair color, placing her in a studio setup with specified camera/lens settings, lighting, makeup, wardrobe, and a pink-to-fuchsia gradient backdrop. The intended outcome is a tightly framed, magazine-ready chest-up image featuring dollar-bill hair curlers and a fan of banknotes, with sharp eye focus, shallow depth of field, refined skin retouching, and polished cinematic color and shine.

betterpromptme Skill

This prompt instructs an image-generation system to recreate a specific fashion model exactly as shown in a provided reference—matching face, hair, body proportions, outfit details, and accessories—while keeping the same studio setup, lighting, background, and camera look, and to output a high-detail, sharp, full-body catalog-style studio photo with a subtle dynamic standing pose and neutral professional expression.

betterpromptme Skill

This prompt instructs an image generation/editing system to produce a high-resolution vertical portrait that preserves the subject’s exact real-world likeness while adding a physically plausible keychain in their hand featuring a small, stylized 3D cartoon version of the same person. The result should look like professional photography—natural skin texture, cinematic lighting, shallow depth of field, warm indoor background—while the mini figure reads as a toy-like 3D character with matching outfit and clear resemblance, integrated with consistent lighting, realistic metallic materials, and correct shadows, avoiding artifacts like distorted hands, mismatched identity, or watermarks.

betterpromptme Skill

Ce prompt demande de générer une image ultra‑réaliste en 4K (16:9) d’un homme dont le visage doit rester parfaitement identique à celui de l’image fournie, intégré dans une scène de temple au trésor façon aventurier‑archéologue. Il précise sa tenue complète, sa pose (remplacer un calice doré par une pierre), l’environnement (portes monumentales, trésors poussiéreux, plantes, racines, stalactites et petits animaux), ainsi que le cadrage, la mise au point, la palette, l’éclairage en clair‑obscur et des contraintes négatives. Le résultat attendu est un plan moyen cinématographique, fortement texturé et contrasté, centré sur son visage concentré et le calice éclairé par un rayon de lumière, sans altérer le visage ni ajouter de nouvelles personnes.

betterpromptme Skill

This prompt sets up an AI feng shui advisor that, given a specific planned activity plus the user’s date and birth details and city, calculates a weighted confidence score and outputs a locale-appropriate recommendation as “should do” vs “should not do” percentages first, followed by a brief rationale and practical tips such as favorable time windows, directions, colors, and alternative better dates, while explicitly noting assumptions when data is missing and keeping the guidance non-harmful and for reference only.

betterpromptme Skill

This prompt directs an image-generation model to create an ultra-photorealistic, dreamy fashion editorial portrait that precisely matches the subject’s facial identity from a provided reference image. It specifies styling and beauty details (oversized shirt, warm natural makeup, porcelain glow), a cinematic rainy-afternoon window scene with raindrops and bokeh reflections, and a very shallow depth of field for a creamy blurred background. The intended outcome is a hyper-detailed, intimate, romantic portrait with warm, soft lighting and a high-end editorial mood.

betterpromptme Skill

This prompt directs an image-generation task to create a high-quality lifestyle photograph of a woman in a modern minimalist living room, specifying her outfit, the neutral curated decor, and the soft natural lighting. The expected result is an Instagram-ready, sharply focused, well-composed scene with a calm, airy atmosphere and elegant, cohesive styling.

betterpromptme Skill

This prompt instructs an AI image generator to create a high-angle, sunlit fashion-style photograph of a casually posed young woman against a distinctive matte pink wall, emphasizing a clean, minimal composition with strong midday shadows and specific wardrobe, accessories, gesture, and nearby phone details to achieve a vibrant California street-photo look.

betterpromptme Skill

This prompt instructs the AI to generate an ultra-high-resolution, minimalist cinematic studio birthday portrait of a woman, precisely matching the subject’s facial features from a reference image, posed crouching in a white studio while blowing out a single candle on a small white cake, styled in an oversized black blazer with sheer tights and bow-accent heels, and composed with sharp dual-spotlight shadows plus celebratory props (red roses and metallic number balloons) to produce a clean, elegant, editorial final photograph.

betterpromptme Skill

This prompt instructs the AI to generate a 3D icon of a specified item, matching the visual style of a provided reference image, and to output the icon with a transparent background. The result is a clean, style-consistent 3D asset suitable for UI, branding, or design use without any background layer.

betterpromptme Skill

This prompt instructs the AI to use a provided reference image of a person to generate a full-body, high-fashion portrait featuring a sage green embroidered abaya, matching hijab, and gold earrings, set against a softly blurred twilight Arabian souq with lantern bokeh. The intended outcome is a cinematic, photorealistic vertical (9:16) image with a razor-sharp subject and creamy background blur, styled as if shot on a 35mm lens at f/1.4.

betterpromptme Skill

This prompt instructs an AI video generator to create a short, photorealistic vertical CCTV-style clip of a suburban front yard viewed from a fixed camera above a front door, with bright natural sunlight and subtle grain. A specified subject approaches along the pathway, stops beneath the camera, looks up, smiles, and says “Hello,” while realistic ambient outdoor sounds and a voice matched to the subject’s characteristics complete the scene.

betterpromptme Skill

This prompt instructs an image model to generate a photorealistic, cinematic 4:5 full-body winter fashion portrait on a snowy alpine slope, keeping the subject’s face, hairstyle, and proportions identical to a provided reference while styling them in a glossy red fur-trimmed puffer jacket with matching ski accessories and crisp natural daylight; the outcome is an 8K editorial-quality image with realistic cold-weather details (breath vapor, wind, snow texture), cool-toned lighting, and a modern luxury-meets-adventure mood.

betterpromptme Skill

This prompt directs an AI tool to generate a manga-style comic using the “Nano Banana Pro” theme, typically featuring banana-inspired characters or motifs, and to output a cohesive set of illustrated panels (or a single comic image) with consistent character design, expressive line art, and readable dialogue or captions that match the intended tone and story beat.

betterpromptme Skill

This prompt instructs an AI photo-editing assistant to remove a user-specified unwanted object from an uploaded image and convincingly inpaint the missing area so the background continues naturally. It emphasizes preserving everything else in the photo—especially the main subject—while matching lighting, shadows, texture, and perspective to avoid visible artifacts. The outcome is a clean, realistic image where the removed item leaves no trace and the scene looks unchanged apart from that deletion.

betterpromptme Skill

This prompt instructs the AI to create a portrait-oriented, fine-art black-and-white outdoor bridal portrait that preserves the subject’s exact facial identity while placing her in an elegant off-shoulder gown with a soft, translucent veil, gently looking down at a bouquet of roses and wildflowers. It specifies a serene, emotional mood with natural daylight, smooth tonal gradients, optional subtle film grain, and a shallow depth of field against softly blurred dark trees, resulting in a timeless, romantic image with soft contrast and no distracting elements or identity-altering stylization.

betterpromptme Skill

This prompt instructs an AI image generator to create a cohesive, three-panel vertical triptych of photorealistic, golden-hour fine-art portraits featuring the same subject consistently across all frames, with a warm pastoral film aesthetic. It specifies the composition and actions for each panel (an intimate restful close-up, a wider full-body meadow scene, and an expressive sensory close-up), recurring props and flora, and detailed camera/film characteristics (lens, shallow depth of field, grain, halation, color grade) to produce a timeless, serene, cinematic sequence that feels like analog film photography.

betterpromptme Skill

This prompt directs an image-generation system to create an ultra-realistic, cinematic Christmas-themed full-body seated portrait based on a provided reference photo, preserving the subject’s exact facial identity, hairstyle, and hair color while applying gender-appropriate festive wardrobe styling. It specifies a cozy holiday environment (decorated tree, gifts, carpet, fairy lights), golden-hour lighting with warm ambient glow, detailed camera settings, a calm serious expression and floor-level perspective, plus controlled post-processing for natural skin texture and subtle festive sparkle. The outcome is a high-end fashion-art holiday portrait that matches the reference person precisely while delivering a warm, dreamy, ultra-detailed seasonal scene.

betterpromptme Skill

This prompt instructs the AI to act as an experienced educator and curriculum designer and produce a complete, teacher-ready lesson plan for a 60-minute class using the provided subject, grade/age level, and duration inputs. The resulting output is a clearly organized plan with labeled sections—overview, measurable learning objectives, required materials, timed introduction/main activities/conclusion, formative and summative assessments, and an optional extension or homework—written in professional, practical language designed for active student engagement and easy classroom reuse.

betterpromptme Skill

This prompt instructs an image model to generate a high-resolution, square 3D render of a chosen object as a floating icon styled like retro N64-era plastic hardware, using rounded semi-matte forms, bold saturated colors, subtle seams, and gentle reflections, centered on a clean white background with no text, shadows, or controller-shaped elements.

betterpromptme Skill

This prompt instructs the AI to write a polished, academically styled essay of a specified length on a user-chosen topic. It requires a clear thesis-led structure with an engaging introduction, evidence-based body paragraphs, and a concise conclusion that synthesizes the main arguments. The outcome is a coherent, grammatically accurate, critically informed essay that incorporates appropriate citations in the requested format where needed.

betterpromptme Skill

This prompt guides an AI to act as an advertising strategist and behavioral psychologist by taking product details, performance goals, market context, and a desired style to build three target personas, generate three ad concepts, run multiple rounds of persona-based critique and iterative refinement, and then select the best-performing ad with clear reasoning and actionable targeting and optimization recommendations.

betterpromptme Skill

This prompt directs an image-generation/editing model to create a realistic, surreal cinematic portrait using the uploaded person’s face, placing them in a dark minimalist room where a projector paints vivid orange-gold, goldfish-like light patterns across the surfaces to evoke an underwater illusion. It specifies gender-dependent wardrobe styling, dreamy immersive lighting with blue haze, soft depth of field, rim lighting, and reflective floor effects, resulting in a calm, contemplative scene that blends futuristic projection technology with magical aquatic fantasy while preserving facial likeness and photographic realism.

betterpromptme Skill

This prompt instructs the AI to act as an experienced business communication specialist and generate a ready-to-send professional email template based on user-supplied details such as email type, recipient, purpose, key points, tone, and context. The result is a clearly structured, concise, and persuasive email—including a short subject line, appropriate greeting, organized body (with bullets when helpful), a specific call-to-action, and a professional closing—formatted for immediate use with customizable placeholders and checked against clarity, tone consistency, and completeness standards.

betterpromptme Skill

This prompt instructs an AI image compositor to create a single 4K photorealistic esports-stage photo by combining a user’s uploaded portrait with a specific reference photo of Faker, preserving both identities exactly while matching lighting, lens characteristics, depth of field, and stage ambience. The expected outcome is a believable live-event style image where Faker stands beside the user with his arm on the user’s shoulder, the user makes a “shhh” gesture toward the camera, and the background features cinematic tournament lighting and crowd/stage details with seamless blending and no visible AI artifacts.

betterpromptme Skill

This prompt instructs an image-generation workflow to take an uploaded portrait and convert it into a Pixar-inspired 3D animated character by preserving the person’s defining facial traits (hair, skin tone, eye shape, and face structure) while enhancing them with stylized cartoon proportions. It specifies the desired look and rendering outcome—soft skin with subtle blush, slightly exaggerated ears and a rounded nose, natural soft lighting, a clean alpha-style background, an ultra-detailed Octane-like cinematic finish, and a vertical 2:3 upper-body composition with a vibrant, whimsical tone.