Jessitoii

Use when writing, modifying, or debugging browser automation code for a specific job application portal — Greenhouse, Lever, Workday, or an unknown custom ATS. Triggers on: "ATS adapter", "new adapter", "write new adapter", "Workday form", "Greenhouse selector", "Lever integration", "unknown portal", "custom ATS", "LLM vision fallback", "DOM selector not found", "form automation", "apply engine", "portal scraper", "shadow DOM", "form is being filled".

Use when implementing or modifying the CV tailoring pipeline — selecting the right CV version, reordering project/skill blocks for a specific job, generating dynamic cover letters, or updating the tailoring prompt contract. Triggers on: "CV seçimi", "CV tailoring", "özgeçmiş optimizasyonu", "cover letter", "immutable block selection", "CV versiyonu", "keyword injection", "tailored summary", "proje sıralama".

Use when writing database migrations, queries, or ORM models for this project. Triggers on: "migration yaz", "tablo oluştur", "PostgreSQL şema", "PGVector", "SQL sorgu", "Alembic", "index ekle", "ENUM güncelle", "application_data JSONB", "cv_performance_stats view", "interaction_logs", "veritabanı modeli".

Use when implementing or debugging the 3-layer duplicate application detection system — DB hash check, email parsing for manual applications, or semantic vector similarity detection. Triggers on: "duplicate prevention", "mükerrer başvuru", "aynı ilana iki kez", "email parsing", "Gmail API kontrolü", "semantic duplicate", "URL hash", "95% similarity", "already applied".

Use when building or debugging the Human-in-the-Loop Telegram integration — pausing agent tasks for user approval, sending inline button messages, handling webhook callbacks, and resuming suspended Celery tasks. Triggers on: "Telegram bot", "kullanıcı onayı", "AWAITING_APPROVAL", "inline button", "webhook callback", "task dondur", "maaş beklentisi sor", "HITL", "async approval", "human-in-the-loop".

Use when building or triggering the automated interview preparation pipeline — detecting interview invitations from email/calendar, running company research, LinkedIn analysis, and generating prep documents. Triggers on: "mülakat daveti", "interview detected", "ResearchAgent", "prep-doc", "şirket araştırma", "LinkedIn profil analizi", "takvim daveti", "ICS", "next steps email", "teknik soru tahmini", "interview prep".

Use when implementing or debugging the LLM provider fallback chain — distinguishing per-minute vs per-day rate limits, routing between Groq/Cerebras/Ollama, or handling invalid JSON responses from LLMs. Triggers on: "rate limit", "429 hatası", "model zinciri", "fallback chain", "Groq limit", "Cerebras", "günlük kota", "dakikalık limit", "AllModelsExhausted", "call_with_fallback", "provider değiştir".

Use when building or modifying the job relevance scoring pipeline — embedding-based local filtering, LLM-based deep scoring, JSON output validation, or the confidence threshold routing logic. Triggers on: "relevance score", "ilan puanlama", "embedding filter", "cosine similarity", "Groq scoring", "LLM analiz", "confidence threshold", "auto_apply karar", "match reasoning", "hard-skill analizi", "vektör benzerliği".

Use this skill when implementing background calculation (calories, cost, macros) for TheMealDB recipes in BOTH the Recipes tab AND the Discover tab — including optimistic UI loading, per-ingredient skeleton states, enabling "Günlüğe Ekle / Add to Log" button before calculation completes, partial log updates after calculation finishes, and local completion notifications. Also use when the Recipes tab already has progressive rendering but is missing the early "Add to Log" option or completion notification.

Use when calling an external API (Migros, TheMealDB, OpenFoodFacts, USDA, MyMemory, Unsplash, Cerebras) or adding a new service in KaloriTabak. Triggers on "Migros'tan fiyat çek", "tarif ara", "kalori getir", "çevir", "API çağrısı".

Use when implementing calorie calculation, macro calculation, recipe cost calculation, BMR/TDEE, unit conversion, or the smart recipe suggestion logic in KaloriTabak. Triggers on "kalori hesapla", "maliyet hesapla", "BMR", "TDEE", "makro", "birim dönüştür", "porsiyon".

Use when building a new reusable UI component for KaloriTabak. Triggers on "component oluştur", "yeni bileşen", "card yap", "buton ekle", "UI component".

Use when adding a new SQLite query, creating a migration, or modifying the database schema in KaloriTabak. Triggers on "add a query", "new migration", "yeni tablo", "veritabanı", "SQLite".

Use when creating a new screen or page in the KaloriTabak app. Triggers on "add a screen", "create a new tab", "new route", "yeni ekran", "ekran oluştur".

Use when adding new state, actions, or selectors to a Zustand store in KaloriTabak. Triggers on "store'a ekle", "state güncelle", "zustand", "global state", "store action".

Guides the implementation of the ReAct loop, orchestrator, memory, and planner. Use this when working on any file inside backend/agent/.

Shared interfaces and contracts between all agents. Every agent MUST read this skill before starting any implementation. This is the single source of truth for all cross-agent boundaries.

Guides FastAPI backend implementation. Use this when working on main.py, routers, or any API endpoint.

Guides implementation of security tools and exploit modules. Use this when working on shell_tools.py, exploit_tools.py, or any offensive security functionality.

Guides the implementation of any tool inside backend/tools/. Use this when creating, editing, or debugging a tool class.

Guides workspace and session management. Use this when working on session_manager.py, workspace_manager.py, or anything related to session isolation.

Enforces a high-fidelity Lyft-inspired design system across all React Native components. Utilizes visual resources and strict design tokens to ensure a professional, minimalist, and high-contrast user experience.

Manages migration from legacy React Navigation patterns to Expo Router (v3+) and enforces a clean, modular project structure.

Scans and refactors the codebase to remove the "Taxi" branding, replacing it with "Route" or "Ride" to ensure compliance with Google Play Store transportation policies.

Performs deep dependency tracing to eliminate monetary, wallet, and transaction-related logic from Prisma schemas and Express routes.

Identifies and eliminates all monetary, wallet, and transaction-related logic from Prisma schemas and Express routes to convert the app into a free peer-to-peer service.

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Reviews context assembly, audit traces, and budget reductions.

Helps retrieve, summarize, and write durable memory carefully.