all publishers

ndpvt-web

@ndpvt-web source repo

659 published skills · page 2 of 7

  1. ▌
    Paperbanana Automating Academic Illustration · ndpvt-web
    Generate publication-ready academic illustrations using a multi-agent pipeline inspired by PaperBanana. Orchestrates retrieval, planning, styling, rendering, and self-critique agents to produce methodology diagrams and statistical plots. Use when: 'create a figure for my paper', 'generate a methodology diagram', 'make a publication-ready illustration', 'draw an architecture diagram for this system', 'create a statistical plot for these results', 'illustrate this pipeline'.
    14 repo stars
  2. ▌
    Refer Agent Collaborative Multi Agent System · ndpvt-web
    Build collaborative multi-agent systems that use alternating reasoning-reflection cycles with specialized agent roles, coarse-to-fine selection, dynamic focus layouts, and chain-of-reflection verification. Use when asked to: 'build a multi-agent pipeline with self-verification', 'create agents that reflect on their own outputs', 'design a coarse-to-fine selection system', 'implement a questioner-responder verification loop', 'build a reasoning pipeline with iterative refinement', 'create a multi-step agent workflow with feedback loops'.
    14 repo stars
  3. ▌
    Safepred Predictive Guardrail Computer Using · ndpvt-web
    Implement predictive safety guardrails for computer-using agents and automated pipelines using world-model-based risk prediction. Catches long-term risks that reactive checks miss by simulating future states before executing actions. Use when: 'add safety guardrails to my agent', 'predict risks before executing shell commands', 'build a predictive safety layer', 'guardrail my automation pipeline', 'prevent irreversible actions in my agent workflow', 'add a world model safety check to my CUA'.
    14 repo stars
  4. ▌
    Stalled Biased Confused Uncovering Reasoning · ndpvt-web
    Systematic root cause analysis for cloud/distributed system failures using a 16-category reasoning failure taxonomy and multi-hop fault propagation tracing. Use when: 'diagnose this production incident', 'find the root cause of this outage', 'trace this failure through our microservices', 'why is this service failing', 'analyze these alerts and logs to find what broke', 'RCA for this distributed system issue'.
    14 repo stars
  5. ▌
    Comprehensive Evaluation Software Engineering · ndpvt-web
    Evaluate and optimize LLM-driven software engineering workflows across five task types (bug fixing, feature development, code refactoring, technical copywriting, research synthesis) using efficiency-aware metrics that go beyond correctness. Detects and eliminates loop inefficiency and inference inefficiency patterns in agentic tool usage. Use when: "evaluate my coding agent workflow", "optimize tool call efficiency", "benchmark LLM SE performance", "reduce agentic coding cost", "audit agent tool usage", "compare efficiency of coding approaches".
    14 repo stars
  6. ▌
    Contextevolve Multi Agent Context Compression · ndpvt-web
    Multi-agent iterative code optimization using context compression. Decomposes optimization into three agents (Summarizer, Navigator, Sampler) that mirror RL state/policy/replay to evolve better code across iterations. Trigger phrases: 'optimize this algorithm iteratively', 'evolve better code', 'multi-agent code optimization', 'compress optimization context', 'iterative code improvement with agents', 'ContextEvolve approach'
    14 repo stars
  7. ▌
    Culturally Grounded Personas Characterization · ndpvt-web
    Generate and evaluate culturally-grounded LLM personas using World Values Survey variables, Inglehart-Welzel Cultural Map positioning, and Moral Foundations Theory profiling. Use this skill when the user asks to: - "create culturally diverse personas for user testing" - "build a persona pipeline grounded in real survey data" - "evaluate cultural bias in LLM-generated personas" - "map synthetic personas onto the Inglehart-Welzel cultural map" - "generate moral profiles for different cultural configurations" - "simulate culturally conditioned survey responses"
    14 repo stars
  8. ▌
    Deltaevolve Accelerating Scientific Discovery · ndpvt-web
    Iteratively evolve code solutions using momentum-driven semantic deltas instead of full-code histories. Use when: 'evolve a better heuristic for bin packing', 'optimize this algorithm iteratively', 'use LLM-driven evolution to improve this function', 'find a better solution through evolutionary search', 'iteratively refine this solver', 'apply DeltaEvolve to discover a better algorithm'.
    14 repo stars
  9. ▌
    Discovering Process Outcome Credit Multi Step · ndpvt-web
    Apply Step-wise Marginal Information Gain (MIG) credit assignment to multi-step reasoning tasks. Evaluates each reasoning step by its marginal contribution toward the correct answer rather than by position or final outcome alone. Use this skill when asked to: "evaluate my chain-of-thought", "score each reasoning step", "find where my logic goes wrong", "credit assign my solution steps", "debug my multi-step reasoning", "identify which steps actually matter in this derivation".
    14 repo stars
  10. ▌
    Domain Specific Knowledge Graphs RAG Enhanced · ndpvt-web
    Build scope-matched knowledge graph RAG pipelines where retrieval precision beats breadth. Constructs domain-specific KGs from scientific literature, selects scope-aligned subgraphs for retrieval, and injects focused context into LLM prompts — avoiding the accuracy loss caused by union-based retrieval. Use when: "Build a knowledge graph RAG pipeline for medical questions", "Add domain-specific retrieval to my LLM app", "My RAG pipeline returns too much irrelevant context", "Help me scope my knowledge graph to match my query domain", "Design a biomedical QA system with knowledge graphs", "Reduce noise in my retrieval-augmented generation system"
    14 repo stars
  11. ▌
    Es Memeval Benchmarking Conversational Agents · ndpvt-web
    Build and evaluate long-term memory systems for conversational agents using the ES-MemEval five-capability framework (information extraction, temporal reasoning, conflict detection, abstention, user modeling). Use when: 'evaluate my chatbot memory', 'build long-term user memory for my agent', 'benchmark conversational memory capabilities', 'add personalization memory to my dialogue system', 'detect contradictions in user history', 'implement temporal reasoning over chat sessions'.
    14 repo stars
  12. ▌
    Harmoni Multimodal Personalization Multi User · ndpvt-web
    Build multi-user personalization pipelines with per-user profile tracking, multimodal perception, and LLM-driven contextual response generation following the HARMONI architecture. Use when: 'build a multi-user chatbot with memory', 'personalize responses per user', 'track user profiles across sessions', 'add speaker identification to my app', 'create a system that remembers each user differently', 'build an assistive robot conversation system'.
    14 repo stars
  13. ▌
    Human Aligned Enhancement Programming Answers · ndpvt-web
    Enhance programming answers by classifying user feedback comments as actionable or non-actionable, then surgically incorporating valid concerns while preserving original intent. Based on the AUTOCOMBAT technique from Bappon et al. (2026). Trigger phrases: "improve this answer based on the comments", "incorporate feedback into this code answer", "fix this Stack Overflow answer using the comments", "enhance this programming answer with user feedback", "refine this technical answer based on reviewer comments", "update this code based on the critique"
    14 repo stars
  14. ▌
    Hyperoffload Graph Driven Hierarchical Memory · ndpvt-web
    Design and implement compiler-driven hierarchical memory offloading for LLM inference and training on multi-tier memory systems. Applies graph-level scheduling of data movement to hide memory transfer latency behind computation. Use when: 'optimize LLM memory offloading', 'schedule prefetch and evict in computation graph', 'hierarchical memory management for inference', 'hide data transfer latency in ML pipeline', 'compiler pass for memory placement', 'reduce peak GPU memory with offloading'.
    14 repo stars
  15. ▌
    Just In Time Reinforcement Learning Continual · ndpvt-web
    Implement JitRL-style continual learning for LLM agents: training-free policy optimization via experience memory, advantage estimation, and logit modulation. Use when asked to 'add experience memory to an agent', 'implement continual learning without fine-tuning', 'build a JitRL agent', 'optimize agent actions from past trajectories', 'add non-parametric RL to an LLM pipeline', or 'make my agent learn from its mistakes at inference time'.
    14 repo stars
  16. ▌
    Learning Irrecoverable Error Localized Policy · ndpvt-web
    Debug multi-step tool-using agent pipelines by localizing the first irrecoverable error via binary-search rollback, then concentrating fixes on that critical step and its downstream suffix. Use when: 'find where my agent pipeline first breaks', 'debug this tool chain failure', 'locate the root cause in my multi-step workflow', 'which step killed this reasoning chain', 'binary search for the breaking step', 'error-localize my agent trajectory'.
    14 repo stars
  17. ▌
    Odysseyarena Benchmarking Long Horizon Active · ndpvt-web
    Design and run inductive agent benchmarks where LLMs must discover hidden rules through long-horizon interaction loops rather than following explicit instructions. Use when the user mentions 'inductive agent evaluation', 'long-horizon benchmarking', 'hidden rule discovery', 'active exploration benchmark', 'OdysseyArena', or 'agent world-model induction'.
    14 repo stars
  18. ▌
    Omnireview Large Scale Benchmark LLM Enhanced · ndpvt-web
    Build reviewer/expert recommendation systems using LLM-generated semantic profiles and Multi-gate Mixture-of-Experts (MMoE) multi-task learning. Applies the OmniReview Pro-MMoE architecture for matching documents to qualified reviewers or experts. Use when: 'recommend reviewers for this paper', 'build an expert matching system', 'find qualified reviewers from a candidate pool', 'rank experts by relevance to a submission', 'build a reviewer recommendation pipeline', 'match papers to domain experts'.
    14 repo stars
  19. ▌
    Openguandan Large Scale Imperfect Information · ndpvt-web
    Build AI agents for the OpenGuanDan imperfect-information card game benchmark. Covers WebSocket client implementation, game state parsing, action selection, and agent strategy design for the GuanDan four-player cooperative-competitive card game. Trigger phrases: 'build a GuanDan agent', 'OpenGuanDan bot', 'imperfect information game AI', 'card game agent', 'GuanDan WebSocket client', 'multi-agent card game'.
    14 repo stars
  20. ▌
    Raicl Retrieval Augmented In Context Learning · ndpvt-web
    Build retrieval-augmented in-context learning (RAICL) pipelines that convert time-series or signal data into images and classify them with vision-language models using dynamically retrieved few-shot examples. Trigger phrases: 'classify EEG with a VLM', 'retrieval-augmented in-context learning', 'convert signals to images for GPT/Gemini', 'few-shot VLM classification', 'RAICL pipeline', 'time-series image classification with LLM'
    14 repo stars
  21. ▌
    Reliable Responsible Foundation Comprehensive · ndpvt-web
    Audit and harden AI/ML systems for reliability and responsibility using an 8-dimension framework covering bias, security, privacy, hallucination, uncertainty, explainability, distribution shift, alignment, and AIGC detection. Use this skill when users say things like: - "Audit my LLM application for safety and fairness" - "Add responsible AI checks to my ML pipeline" - "How do I detect hallucinations in my model output?" - "Harden my AI system against adversarial attacks" - "Add bias detection to my NLP pipeline" - "Implement watermarking for AI-generated content"
    14 repo stars
  22. ▌
    Robustexplain Evaluating Robustness LLM Based · ndpvt-web
    Evaluate robustness of LLM-generated recommendation explanations under realistic user behavior noise. Use when: 'test explanation robustness', 'evaluate recommender explanation stability', 'perturb user history and measure explanation drift', 'benchmark LLM explanation consistency', 'build a robustness evaluation pipeline for recommendation agents', 'stress-test explanation generation under noisy inputs'.
    14 repo stars
  23. ▌
    Supchain Bench Benchmarking Real World Supply · ndpvt-web
    Build reliable long-horizon supply chain agents using the SupChain-ReAct pattern: multi-path ReAct trajectories with majority voting for autonomous tool orchestration without handcrafted SOPs. Use when asked to 'build a supply chain agent', 'orchestrate multi-step tool calls for order management', 'diagnose fulfillment issues', 'create an SOP-free agent workflow', 'implement long-horizon tool calling', or 'build an e-commerce order diagnostic system'.
    14 repo stars
  24. ▌
    Tracecoder Trace Driven Multi Agent Framework · ndpvt-web
    Trace-driven debugging framework for LLM-generated code. Uses diagnostic probe instrumentation, causal trace analysis, and historical lesson learning to iteratively fix buggy code. Triggers: 'debug this code with traces', 'instrument and fix this function', 'trace-driven debugging', 'find the root cause of this bug', 'iteratively repair this code', 'why does this code fail these tests'
    14 repo stars
  25. ▌
    Tsrbench Comprehensive Multi Task Multi Modal · ndpvt-web
    Evaluate and build multi-modal time series reasoning pipelines using the TSRBench framework. Covers perception, reasoning, prediction, and decision-making over time series data represented as text, plots, or both. Use when: 'evaluate LLM on time series tasks', 'build a time series reasoning benchmark', 'test model on anomaly detection and forecasting', 'analyze time series with multiple modalities', 'compare text vs visual time series performance', 'create time series QA evaluation'.
    14 repo stars
  26. ▌
    Agent2agent Threats Safety Critical Assistants · ndpvt-web
    Threat model multi-agent LLM systems using the AgentHeLLM framework -- formally separating asset identification from attack path analysis with graph-based poison/trigger path discovery. Use when: 'threat model my agent system', 'find attack paths in my A2A architecture', 'analyze security of my LLM agents', 'map attack surfaces for my multi-agent app', 'identify poison paths in my agent graph', 'what are the safety risks in my agent pipeline'.
    14 repo stars
  27. ▌
    Autoregressive Yet Revisable Decoding Revision · ndpvt-web
    Generate secure code using Stream of Revision — an in-decoding self-correction technique that backtracks and patches vulnerable code spans during generation rather than after it. Trigger phrases: - "generate secure code" - "fix security vulnerabilities in this code" - "write safe C/C++ code" - "review and revise code for security" - "backtrack and fix this vulnerability" - "self-correcting code generation"
    14 repo stars
  28. ▌
    Benchmarking Uncertainty Calibration Long Form · ndpvt-web
    Implement uncertainty quantification and calibration assessment for LLM-generated long-form answers. Apply answer-frequency consistency, verbalized confidence elicitation, token-level analysis, and multi-metric calibration benchmarking based on the UQ framework from Müller et al. (2026). Trigger phrases: - "measure how confident the model is in this answer" - "calibrate uncertainty on these QA results" - "benchmark uncertainty quantification for my LLM pipeline" - "which uncertainty method should I use for scientific QA" - "detect unreliable LLM answers" - "evaluate calibration of model confidence scores"
    14 repo stars
  29. ▌
    Chart Specification Structural Representations · ndpvt-web
    Generate high-fidelity plotting code from chart images or descriptions using structured intermediate specifications. Decomposes charts into semantic topology (type, coordinates, domains, series) and runtime numerical facts before producing code, preventing hallucinated data and structural errors. Trigger phrases: - "convert this chart image to code" - "recreate this plot in matplotlib" - "generate plotting code from this chart" - "reproduce this visualization programmatically" - "write code that matches this chart exactly" - "chart to code"
    14 repo stars
  30. ▌
    Clustering Driven Memory Compression On Device · ndpvt-web
    Compress user-specific memories for LLM personalization by clustering semantically similar memories and merging within clusters, reducing token count while preserving generation quality. Based on Bohdal et al. (ICASSP 2026). Use this skill when the user mentions: - "compress memories for context window" - "reduce memory tokens for on-device LLM" - "cluster and merge user memories" - "personalization with limited context budget" - "memory-efficient prompt construction" - "on-device LLM memory management"
    14 repo stars
  31. ▌
    Davinci Dev Agent Native Mid Training Software · ndpvt-web
    Apply daVinci-Dev's agent-native workflow to software engineering tasks: navigate repos, localize bugs, plan edits, apply structured patches, and verify with tests. Use when asked to 'fix this issue in the repo', 'resolve this bug across files', 'apply agentic SWE workflow', 'navigate and patch this codebase', 'use agent-native approach to solve this', or 'debug and fix with test verification'.
    14 repo stars
  32. ▌
    Deepimagesearch Benchmarking Multimodal Agents · ndpvt-web
    Build agentic image retrieval systems that perform multi-step contextual reasoning over visual histories instead of isolated semantic matching. Use when: 'build a context-aware image search agent', 'retrieve images using temporal reasoning', 'search photos by contextual clues across events', 'implement dual-memory agent for image retrieval', 'create a visual history exploration pipeline', 'benchmark multimodal agents on retrieval tasks'.
    14 repo stars
  33. ▌
    Deepplanning Benchmarking Long Horizon Agentic · ndpvt-web
    Solve long-horizon planning tasks with verifiable constraints using the DeepPlanning methodology: proactive information gathering, local constraint reasoning, and global constrained optimization. Use when asked to 'plan a multi-day trip with budget constraints', 'build a shopping optimizer with coupons and sizing', 'create an agent that handles complex multi-step planning', 'design a constraint-satisfaction planner', 'optimize across interdependent decisions with budgets', or 'build a planning benchmark with verifiable solutions'.
    14 repo stars
  34. ▌
    Dep Search Learning Dependency Aware Reasoning · ndpvt-web
    Dependency-aware multi-step reasoning with persistent memory for complex questions requiring information retrieval across multiple sources. Use when: 'answer this multi-hop question', 'research this topic step by step', 'find information that depends on other lookups', 'break down this complex question', 'trace the reasoning chain for this query', 'search and synthesize across multiple sources'.
    14 repo stars
  35. ▌
    Eft Cot Multi Agent Chain Of Thought Framework · ndpvt-web
    Build multi-agent emotion-focused therapy (EFT) reasoning pipelines for empathetic mental health Q&A systems. Uses a bottom-up three-stage chain-of-thought: Embodied Perception, Cognitive Exploration, and Narrative Intervention with eight specialized agents. Trigger phrases: 'build an EFT chatbot', 'emotion-focused therapy agent', 'empathetic counseling system', 'multi-agent mental health pipeline', 'somatic-aware therapy bot', 'EFT-CoT reasoning chain'.
    14 repo stars
  36. ▌
    Extracting Recurring Vulnerabilities Black Box · ndpvt-web
    Predict and prevent recurring vulnerabilities in LLM-generated code using the FSTab (Feature-Security Table) technique. Maps observable frontend features to likely backend vulnerabilities based on which LLM generated the code. Use this skill when: - "Audit this LLM-generated app for security vulnerabilities" - "What vulnerabilities does this generated code likely have?" - "Check this AI-generated backend for recurring security issues" - "Build an FSTab mapping for this codebase" - "Run a black-box vulnerability prediction on this web app" - "Assess vulnerability persistence in code from GPT/Claude/Gemini"
    14 repo stars
  37. ▌
    Fademem Biologically Inspired Forgetting Agent · ndpvt-web
    Implement biologically-inspired forgetting mechanisms for LLM agent memory systems. Build dual-layer memory hierarchies with adaptive exponential decay, semantic relevance scoring, and LLM-guided conflict resolution to keep agent context lean and high-quality. Use when: "add forgetting to my agent memory", "implement memory decay for my chatbot", "build an agent memory system with selective retention", "reduce memory bloat in my AI agent", "implement FadeMem-style memory management", "add adaptive memory consolidation to my agent".
    14 repo stars
  38. ▌
    Masalbench Benchmark Contextual Cross Cultural · ndpvt-web
    Build cross-cultural figurative language benchmarks and evaluation pipelines for LLMs. Applies the MasalBench methodology to test whether models truly understand proverbs, idioms, and culturally-embedded expressions -- not just pattern-match surface text. Trigger phrases: 'evaluate LLM cultural understanding', 'benchmark figurative language', 'test proverb comprehension', 'cross-cultural NLP evaluation', 'build idiom benchmark', 'low-resource language evaluation'
    14 repo stars
  39. ▌
    Reflect Transparent Principle Guided Reasoning · ndpvt-web
    Apply the REFLECT constitutional alignment framework to enforce user-defined principles on LLM outputs through a multi-stage pipeline: constitution-conditioned generation, self-evaluation with Likert scoring, self-critique, and principled revision. Use this skill when asked to: "align output to these principles", "apply constitutional review to this response", "reflect on whether this follows our guidelines", "enforce these coding standards on generated code", "review this output against our policy", "run a principled critique and revision pass".
    14 repo stars
  40. ▌
    Sparc RAG Adaptive Sequential Parallel Scaling · ndpvt-web
    Implement multi-agent RAG systems with coordinated sequential-parallel scaling and shared context management for complex multi-hop question answering. Use when: 'build a multi-hop RAG pipeline', 'implement adaptive retrieval with parallel branching', 'create a RAG system that scales retrieval depth and width', 'build a multi-agent retrieval framework', 'implement SPARC-RAG pattern for complex QA', 'design a RAG system with context management across retrieval rounds'.
    14 repo stars
  41. ▌
    Table As Search Formulate Long Horizon Agentic · ndpvt-web
    Structured table-completion framework for long-horizon information seeking. Converts complex research queries into database tables where rows are candidates and columns are constraints/attributes, then orchestrates deep and wide search agents to fill cells systematically. Use when: 'research and compare multiple options', 'find all X that satisfy Y constraints', 'deep dive investigation across sources', 'build a comparison table from web research', 'search for entities matching criteria', 'comprehensive multi-step information gathering'.
    14 repo stars
  42. ▌
    Veri Sure Contract Aware Multi Agent Framework · ndpvt-web
    Generate functionally correct RTL/Verilog code using a contract-aware multi-agent workflow with formal verification. Triggers: 'generate Verilog module', 'write RTL with verification', 'design hardware with formal proof', 'create verified Verilog', 'contract-driven RTL generation', 'fix Verilog with dependency slicing'
    14 repo stars
  43. ▌
    Agentcpm Report Interleaving Drafting Deepening · ndpvt-web
    Generate deep research reports by interleaving evidence-based drafting with reasoning-driven deepening. Uses the WARP (Writing As Reasoning Policy) framework from AgentCPM-Report to dynamically evolve outlines during writing instead of rigidly following a static plan. Trigger phrases: "deep research report", "write a comprehensive analysis", "investigate and write up", "research report on", "deep dive report", "analyze this topic thoroughly and produce a report"
    14 repo stars
  44. ▌
    Automating Computational Reproducibility Social · ndpvt-web
    Diagnose and repair failing computational research code to restore reproducibility. Uses an agent-based iterative workflow: inspect files, identify failures (missing packages, broken paths, version conflicts, missing logic), apply targeted fixes, and rerun in isolated environments. Trigger phrases: 'reproduce this analysis', 'fix this R script', 'make this code reproducible', 'debug this research pipeline', 'repair computational workflow', 'rerun this study'
    14 repo stars
  45. ▌
    Benchmarking Text To Python Against Text To SQL · ndpvt-web
    Generate correct Python/Pandas code from natural language questions over tabular data, applying the Logic Completion Framework (LCF) to resolve ambiguities that SQL handles implicitly. Use when: 'query this CSV with Python', 'convert this SQL to pandas', 'answer this question from the database using Python', 'write pandas code for this data question', 'translate SQL query to Python', 'analyze this table and answer questions about it'.
    14 repo stars
  46. ▌
    Canonical Intermediate Representation LLM Based · ndpvt-web
    Translate natural language optimization problems into executable solver code using a Canonical Intermediate Representation (CIR) schema and multi-agent R2C pipeline. Decomposes operational rules into constraint archetypes and modeling paradigms before generating code. Triggers: "formulate this optimization problem", "write a solver for this scheduling problem", "convert these business rules to constraints", "model this linear program from the description", "generate Gurobi/PuLP code for this OR problem", "help me formulate these operational constraints mathematically".
    14 repo stars
  47. ▌
    Evermembench Benchmarking Long Term Interactive · ndpvt-web
    Build and evaluate long-term conversational memory systems for multi-party, multi-topic dialogues. Implements the EverMemBench framework for stress-testing memory architectures against realistic workplace conversation patterns with temporal evolution, cross-topic interleaving, and role-specific personas. Use when: 'build a memory system for multi-user chat', 'evaluate my RAG memory pipeline', 'benchmark long-term conversation recall', 'test memory across multi-party dialogues', 'design a temporal memory store for chat agents', 'audit retrieval quality for conversational AI'.
    14 repo stars
  48. ▌
    Interpreting Controlling Behavior Constitutions · ndpvt-web
    Learn and apply natural-language constitutions that map prompt edits to predictable model behavior changes. Use atomic concept edits (ACEs) to systematically probe, interpret, and steer LLM or text-to-image model outputs. Trigger phrases: 'learn a constitution for this model', 'find what prompt changes affect output', 'build an ACE constitution', 'probe model behavior with edits', 'steer model output systematically', 'interpret model sensitivity to prompt changes'.
    14 repo stars
  49. ▌
    Multi Agent End To End Vulnerability Management · ndpvt-web
    Detect, confirm, repair, and validate recurring software vulnerabilities using a multi-agent pipeline modeled on MAVM. Builds a vulnerability knowledge base from historical CVE/patch data, then coordinates specialized agents for detection, confirmation, repair, and validation. Trigger phrases: 'find recurring vulnerabilities', 'scan for known vulnerability patterns', 'port security patches across repos', 'detect unpatched code clones', 'end-to-end vulnerability management', 'fix recurring security issues'.
    14 repo stars
  50. ▌
    Pcbschemagen Constraint Guided Schematic Design · ndpvt-web
    Generate PCB schematics from natural language using constraint-guided LLM code generation with knowledge-graph verification. Use when the user says 'generate a PCB schematic', 'design a circuit board', 'create a KiCad schematic from description', 'convert circuit requirements to netlist', 'automate schematic design', or 'generate SKiDL code for a circuit'.
    14 repo stars
  51. ▌
    3 Secbench Large Scale Evaluation Suite Security · ndpvt-web
    Evaluate and harden LLM-based autonomous agents against adversarial attacks using the α³-SecBench layered security framework. Assesses security (attack detection, CWE attribution), resilience (safe degradation), and trust (policy-compliant tool usage) across 7 autonomy layers. Use when: 'audit my LLM agent for security', 'add adversarial resilience to my autonomous system', 'evaluate agent trust and tool safety', 'harden my AI agent against prompt injection', 'security benchmark my LLM pipeline', 'test my agent for hallucinated tool calls'.
    14 repo stars
  52. ▌
    Constitutional Spec Driven Development Enforcing · ndpvt-web
    Enforce security by construction in AI-generated code using Constitutional Spec-Driven Development (CSDD). Creates a versioned security constitution document mapping CWE/OWASP vulnerabilities to enforceable constraints, then generates code that satisfies those constraints with full traceability. Triggers: 'generate secure code', 'create a security constitution', 'build with security by construction', 'CSDD for my project', 'enforce CWE constraints', 'secure code generation with traceability'
    14 repo stars
  53. ▌
    Diffusion Pretrained Dense Contextual Embeddings · ndpvt-web
    Build production retrieval systems using pplx-embed, diffusion-pretrained dense and contextualized embedding models with INT8 quantization, late chunking for long documents, and multi-stage contrastive training. Use when: 'build a semantic search pipeline', 'set up document retrieval with contextual embeddings', 'implement late chunking for long documents', 'create a multilingual search index', 'optimize embedding storage with quantization', 'add contextualized passage retrieval to RAG'.
    14 repo stars
  54. ▌
    Knowledge Restoration Driven Prompt Optimization · ndpvt-web
    Iteratively optimize LLM prompts for information extraction tasks using self-evaluation feedback loops. Applies the KRPO framework: extract structured data, restore it to natural language, score semantic consistency via NLI, then generate textual gradients to refine the prompt. Includes relation canonicalization to deduplicate and normalize extracted schemas. Trigger phrases: - "Extract relations from text and optimize the prompt" - "Build a self-improving extraction pipeline" - "Optimize my prompt for triplet extraction" - "Extract knowledge graph triples from unstructured text" - "Set up iterative prompt refinement with feedback" - "Canonicalize extracted relations across documents"
    14 repo stars
  55. ▌
    Large Scale Multidimensional Knowledge Profiling · ndpvt-web
    Build multidimensional profiling pipelines for large scientific paper corpora. Combines BERTopic clustering, LLM-structured extraction, and weighted semantic retrieval to analyze research trends, topic lifecycles, dataset/model adoption, and methodological shifts. Use when: 'profile these research papers', 'analyze trends in this paper corpus', 'build a scientific literature pipeline', 'extract structured knowledge from papers', 'track topic evolution across conferences', 'find emerging research directions'.
    14 repo stars
  56. ▌
    Mascot Multi Agent Socio Collaborative Companion · ndpvt-web
    Design and orchestrate multi-agent companion systems where each agent maintains a distinct persona and contributes diverse, non-redundant dialogue. Uses MASCOT's bi-level optimization: persona-level behavioral alignment + group-level collaborative dialogue optimization. Trigger phrases: 'build a multi-agent support system', 'create companion agents with distinct personas', 'design a multi-perspective dialogue system', 'prevent persona collapse in multi-agent chat', 'build a team of AI agents that don't echo each other', 'orchestrate collaborative multi-agent dialogue'.
    14 repo stars
  57. ▌
    Mrag Benchmarking Retrieval Augmented Generation · ndpvt-web
    Build and evaluate biomedical RAG pipelines using the MRAG benchmark methodology. Configures retrieval, prompting, and generation components for medical QA systems. Use when: 'build a medical RAG pipeline', 'evaluate my biomedical QA system', 'optimize retrieval for clinical questions', 'set up PubMed-based RAG', 'benchmark RAG on medical datasets', 'configure RAG for drug interaction extraction'.
    14 repo stars
  58. ▌
    Proopf Benchmarking Improving Professional Grade · ndpvt-web
    Translate natural-language power system operational requirements into executable Optimal Power Flow (OPF) optimization code using differential modification of a canonical formulation. Use when the user asks to 'generate OPF code from a description', 'model power flow constraints', 'add security constraints to OPF', 'translate dispatch requirements to optimization', 'modify OPF parameters from scenario description', or 'build power system optimization from natural language'.
    14 repo stars
  59. ▌
    Sogptspotter Detecting Chatgpt Generated Answers · ndpvt-web
    Detect AI-generated answers in Q&A content using Siamese embedding comparison with reference-answer anchoring. Trigger phrases: 'detect AI-generated answers', 'check if this answer is ChatGPT', 'flag LLM-generated Stack Overflow posts', 'is this answer written by AI', 'AI content detection for technical Q&A', 'screen answers for AI generation'
    14 repo stars
  60. ▌
    Detecting Correcting Hallucinations LLM Generated · ndpvt-web
    Detect and auto-correct hallucinated API calls in LLM-generated Python code using deterministic AST analysis and library introspection. Validates function signatures, parameter names, method existence, and identifier references against a dynamically-built Knowledge Base of real API specs. Use when: "check this generated code for hallucinated APIs", "validate these API calls are real", "fix the fake parameters in this code", "verify this code against actual library signatures", "detect hallucinations in this Python snippet", "auto-correct wrong API usage in generated code".
    14 repo stars
  61. ▌
    Environment In The Loop Rethinking Code Migration · ndpvt-web
    Perform code migrations (dependency upgrades, API changes, framework transitions) with integrated environment verification. Instead of migrating code then hoping it builds, this skill builds and tests inside a real environment at every step, using feedback loops to fix both code and configuration issues. Use when: 'migrate this project from X to Y', 'upgrade dependency version', 'port this codebase to a new framework', 'fix build after dependency update', 'help me upgrade NumPy/React/Django/Spring', 'automate this library migration'.
    14 repo stars
  62. ▌
    Evaluating Retrievalaugmented Generation Variants · ndpvt-web
    Build production-grade natural language to SQL/API pipelines using RAG variant selection (Standard RAG, Self-RAG, CoRAG). Implements iterative query decomposition, hybrid documentation retrieval, and dynamic task classification for enterprise NL interfaces. Trigger phrases: - "Build a natural language to SQL interface with RAG" - "Generate API calls from user questions using retrieval augmented generation" - "Set up a hybrid SQL and REST API generation pipeline" - "Implement CoRAG for enterprise query generation" - "Create a text-to-SQL system with document retrieval" - "Design a retrieval pipeline that handles both database queries and API calls"
    14 repo stars
  63. ▌
    Fat Cat Document Driven Metacognitive Multi Agent · ndpvt-web
    Implement the Fat-Cat document-driven metacognitive agent architecture for complex multi-step reasoning tasks. Uses Markdown documents as global state instead of JSON, a four-stage reasoning pipeline (metacognitive analysis, strategy selection, step decomposition, execution), textual strategy evolution for accumulating task-solving knowledge, and a closed-loop watcher to prevent hallucinations and infinite loops. Trigger phrases: "use fat-cat for this task", "document-driven agent", "metacognitive reasoning pipeline", "markdown state management", "multi-agent with strategy evolution", "fat-cat agent workflow"
    14 repo stars
  64. ▌
    Harnessing Precision Querying Retrieval Augmented · ndpvt-web
    LLM-driven precision querying of structured tabular data via Python/Pandas code generation and retrieval-augmented extraction from unstructured clinical text. Use when: 'query this table in natural language', 'extract information from clinical notes', 'build a RAG pipeline for medical records', 'generate Pandas code from a question', 'answer questions about EHR data', 'set up evaluation for Q&A over datasets'.
    14 repo stars
  65. ▌
    Lhaw Controllable Underspecification Long Horizon · ndpvt-web
    Detect and handle ambiguity in long-horizon agent tasks using the LHAW framework. Systematically identify underspecified Goals, Constraints, Inputs, and Context in task prompts, classify ambiguity severity, and decide when to clarify vs. proceed. Use when: 'analyze this task for ambiguity', 'what's missing from this spec', 'generate underspecified variants', 'test agent robustness to vague instructions', 'find what could go wrong with this prompt', 'stress-test this workflow for unclear requirements'.
    14 repo stars
  66. ▌
    Livemedbench Contamination Free Medical Benchmark · ndpvt-web
    Build contamination-free LLM evaluation pipelines with multi-agent data curation and automated rubric-based scoring. Uses LiveMedBench's three-agent curation framework and bipolar rubric evaluation to assess LLM outputs against granular, case-specific criteria. Trigger phrases: 'build a contamination-free benchmark', 'evaluate LLM with rubrics', 'curate clinical test data', 'automated rubric evaluation pipeline', 'detect data contamination in LLMs', 'multi-agent data curation framework'.
    14 repo stars
  67. ▌
    Ontology To Tools Compilation Executable Semantic · ndpvt-web
    Compile domain ontologies (OWL/RDFS/JSON-LD schemas) into executable tool interfaces with embedded semantic constraints, so LLM agents enforce domain rules during generation rather than post-hoc. Use when: 'compile my ontology into tools', 'enforce schema constraints in agent tools', 'generate MCP tools from OWL', 'build knowledge graph extraction pipeline', 'ontology-driven tool generation', 'semantic constraint enforcement for agents'.
    14 repo stars
  68. ▌
    Unveiling Cognitive Compass Theory Of Mind Guided · ndpvt-web
    Apply Theory-of-Mind (ToM) guided reasoning chains to multimodal emotion analysis tasks. Decomposes emotional reasoning into hierarchical cognitive levels—perception, understanding, and causal cognition—tracking mental states explicitly before reaching conclusions. Use when: 'analyze the emotions in this image/video', 'why does this person feel that way', 'build an emotion reasoning pipeline', 'detect sarcasm or humor in multimodal content', 'evaluate emotional understanding in my model', 'add ToM-based reasoning to my MLLM'.
    14 repo stars
  69. ▌
    Anonymization Enhanced Privacy Protection Mobile G · ndpvt-web
    Implement available-but-invisible privacy protection for mobile GUI agents using PII-aware anonymization with deterministic, type-preserving placeholders. Use when: 'anonymize PII in UI automation', 'build privacy layer for mobile agent', 'protect sensitive data in screenshots', 'add PII detection to Android agent', 'type-preserving placeholder system', 'privacy-safe GUI agent pipeline'.
    14 repo stars
  70. ▌
    Avenir Web Human Experience Imitating Multimodal W · ndpvt-web
    Build robust web automation agents using Mixture of Grounding Experts, experience-imitation planning, and task-tracking checklists. Use when: 'build a web agent', 'automate browser tasks with grounding', 'create a web scraping agent with memory', 'implement element grounding for web automation', 'build a multi-step web task agent', 'add procedural knowledge to a browser agent'.
    14 repo stars
  71. ▌
    Batcoder Self Supervised Bidirectional Code Docume · ndpvt-web
    Apply BatCoder's back-translation technique to improve code and documentation quality bidirectionally. Generate documentation from code, then verify it by reconstructing code from that documentation -- using cycle consistency as a quality signal. Triggers: 'generate documentation from code', 'back-translate code and docs', 'verify documentation quality', 'improve code with back-translation', 'self-supervised code documentation', 'cycle-consistent code generation'
    14 repo stars
  72. ▌
    Chunking Retrieval Re Ranking Empirical Evaluation · ndpvt-web
    Build and optimize two-stage RAG pipelines with bi-encoder retrieval, cross-encoder re-ranking, and empirically-validated chunking strategies. Use when: 'build a RAG pipeline', 'add re-ranking to retrieval', 'optimize chunking for documents', 'set up document QA with re-ranking', 'improve RAG faithfulness', 'two-stage retrieval pipeline'.
    14 repo stars
  73. ▌
    Constructing Multi Label Hierarchical Classificati · ndpvt-web
    Build multi-label hierarchical classifiers for MITRE ATT&CK text tagging using stage-wise classical ML (SGD-SVM + TF-IDF). Use when: 'tag CTI text with ATT&CK', 'classify threat reports with MITRE tactics', 'build hierarchical cybersecurity classifier', 'map CVE descriptions to ATT&CK techniques', 'automate MITRE tagging pipeline', 'multi-label threat classification'.
    14 repo stars
  74. ▌
    Context Augmented Code Generation Programming Know · ndpvt-web
    Enhance code generation with Programming Knowledge Graph (PKG) retrieval, tree pruning, and re-ranking. Uses fine-grained knowledge graph nodes (function blocks, code snippets, documentation paths) to retrieve precisely relevant context, then prunes and re-ranks results to minimize hallucination. Trigger phrases: - "Generate code using knowledge graph context" - "Help me solve this coding problem with relevant examples" - "Retrieve similar code patterns for this function" - "Augment my code generation with external knowledge" - "Find relevant code context and generate a solution" - "Use PKG-style retrieval for this programming task"
    14 repo stars
  75. ▌
    Conversation Non Verifiable Learning Self Evolving · ndpvt-web
    Implements the CoNL (Conversation for Non-verifiable Learning) multi-agent self-play framework for iteratively improving outputs on tasks without clear right answers -- creative writing, code review, API design, UX copy, ethical reasoning, and architectural decisions. Uses structured propose-critique-revise conversations with diagnostic reward scoring to surface the highest-quality solutions. Trigger phrases: - "Improve this through self-critique" or "iteratively refine this" - "Use multi-agent critique to improve my code/writing/design" - "Run a propose-critique-revise loop on this" - "Help me evaluate and improve this where there's no single right answer" - "Self-play review this design decision" - "Meta-evaluate these critiques"
    14 repo stars
  76. ▌
    Corefine Confidence Guided Self Refinement Adaptiv · ndpvt-web
    Confidence-guided self-refinement for adaptive reasoning. Implements the CoRefine pattern: assess confidence in each reasoning step, then decide whether to halt, re-examine, or try a different approach -- reducing wasted compute while maintaining accuracy. Use when: 'solve this step by step with self-correction', 'refine your reasoning until confident', 'adaptively debug this problem', 'use confidence-guided refinement', 'self-correct with backtracking', 'try different approaches if stuck'.
    14 repo stars
  77. ▌
    Cowork X Experience Optimized Co Evolution Multi A · ndpvt-web
    Build multi-agent collaboration systems with experience-driven co-evolution using HTN skill libraries and post-episode optimization. Use when: 'build a multi-agent system that improves over episodes', 'create agents that coordinate in real-time with low latency', 'design a skill library for collaborative agents', 'implement experience-based co-evolution for agent teams', 'optimize multi-agent token budget while improving performance', 'set up HTN-based task decomposition for cooperative agents'.
    14 repo stars
  78. ▌
    Ctrlcot Dual Granularity Chain Of Thought Compress · ndpvt-web
    Compress chain-of-thought reasoning using CtrlCoT's dual-granularity framework: hierarchical semantic abstraction combined with logic-preserving token pruning. Use when asked to 'compress reasoning', 'shorten chain of thought', 'optimize CoT tokens', 'reduce reasoning verbosity', 'budget-controlled reasoning', or 'efficient step-by-step thinking'.
    14 repo stars
  79. ▌
    Event Vstream Event Driven Real Time Understanding · ndpvt-web
    Build event-driven video stream processing pipelines that detect meaningful state transitions instead of processing every frame. Use when asked to: 'build a real-time video understanding system', 'detect events in a video stream', 'process long video with memory', 'reduce redundant frame processing', 'stream video to LLM efficiently', 'build an event-aware video pipeline'.
    14 repo stars
  80. ▌
    Evocodebench Human Performance Benchmark Self Evol · ndpvt-web
    Self-evolving code generation with iterative reflection and revision. Applies a feedback-driven loop where code is submitted, judged, analyzed for failures, and rewritten up to 3 times — tracking correctness, runtime, memory, and algorithmic improvement at each iteration. Use when: 'solve this coding problem and optimize it', 'iteratively improve this solution', 'refine my code until it passes all tests', 'benchmark my solution against human performance', 'reduce the time complexity of this code', 'fix and re-attempt this failing solution'.
    14 repo stars
  81. ▌
    Experience Driven Multi Agent Systems Training Fre · ndpvt-web
    Build self-evolving multi-agent systems that accumulate tool-level expertise through structured interaction without model fine-tuning. Uses GeoEvolver's architecture: retrieval-augmented orchestration, parallel sub-goal exploration, contrastive memory distillation, and root-cause failure attribution. Triggers: 'build a self-evolving agent pipeline', 'create an experience-driven multi-agent system', 'add memory to my agent workflow', 'implement tool exploration with failure learning', 'make agents learn from execution history', 'build a GeoEvolver-style system'
    14 repo stars
  82. ▌
    Graph Anchored Knowledge Indexing Retrieval Augmen · ndpvt-web
    Build iterative RAG pipelines that construct evolving knowledge graphs to anchor retrieval across multiple hops. Use when user says 'multi-hop QA', 'graph-guided retrieval', 'iterative RAG', 'knowledge graph indexing', 'connect evidence across documents', or 'structured retrieval pipeline'.
    14 repo stars
  83. ▌
    Halluverse M3 Multitask Multilingual Benchmark Hal · ndpvt-web
    Detect and classify hallucinations in LLM outputs across languages using the HalluVerse-M3 fine-grained taxonomy (entity-level, relation-level, sentence-level). Trigger phrases: 'check for hallucinations', 'detect hallucinations in this output', 'verify factual consistency', 'multilingual hallucination audit', 'classify hallucination type', 'hallucination detection benchmark'.
    14 repo stars
  84. ▌
    Hybrid Supervised LLM Pipeline Actionable Suggesti · ndpvt-web
    Build hybrid classifier-then-LLM pipelines to extract actionable suggestions from unstructured customer reviews. Use when the user says 'extract suggestions from reviews', 'mine actionable feedback', 'analyze customer complaints for improvements', 'build a suggestion extraction pipeline', 'classify and cluster review feedback', or 'summarize actionable insights from user reviews'.
    14 repo stars
  85. ▌
    Isd Agent Bench Comprehensive Benchmark Evaluating · ndpvt-web
    Build and evaluate LLM-based Instructional Design agents using the ADDIE framework, Context Matrix scenario generation, and multi-judge evaluation. Triggers: 'design a course using ADDIE', 'build an instructional design agent', 'evaluate my ISD pipeline', 'create training program with learning objectives', 'benchmark educational content generation', 'generate instructional scenarios with context variables'
    14 repo stars
  86. ▌
    Lingxidiagbench Multi Agent Framework Benchmarking · ndpvt-web
    Build multi-agent benchmarking systems with role-separated agents (simulator, interviewer, evaluator) for structured multi-turn dialogue evaluation. Inspired by the LingxiDiagBench psychiatric consultation framework. Trigger phrases: 'build a multi-agent benchmark', 'create a doctor-patient simulation', 'evaluate LLMs on multi-turn dialogue', 'benchmark diagnostic reasoning', 'multi-agent consultation framework', 'simulate structured interviews with LLMs'
    14 repo stars
  87. ▌
    Llama 31 Foundationai Securityllm Reasoning 8b Tec · ndpvt-web
    Apply Foundation-Sec-8B-Reasoning cybersecurity reasoning patterns: structured <think> chain-of-thought for CVE-to-CWE mapping, MITRE ATT&CK classification, CVSS scoring, threat intelligence analysis, and multi-hop vulnerability reasoning. Use when the user asks to "analyze a CVE", "map vulnerabilities to CWE", "classify attack techniques", "reason about security threats", "triage a vulnerability", or "build a cybersecurity reasoning pipeline".
    14 repo stars
  88. ▌
    Marti Mars2 Scaling Multi Agent Self Search Reinfo · ndpvt-web
    Multi-agent tree-search code generation using heterogeneous agent collaboration with error-feedback refinement. Spawns multiple agents with distinct strategies to explore solution trees, iteratively refine code via structured error diagnostics, and select the best solution through diversity-aware ranking. Use when: 'solve this hard coding problem with multiple agents', 'use multi-agent search to generate code', 'use MARTI-MARS2 to solve this', 'spawn a team to tackle this algorithm problem', 'use tree search to find the best solution', 'generate code with agent collaboration'.
    14 repo stars
  89. ▌
    Optimizing Small Sample Experience Learning LLM Ba · ndpvt-web
    Implement the ExperienceWeaver hierarchical experience-learning framework to improve text quality from small feedback sets. Distills noisy corrections into structured Tips and Strategies, then injects them into a multi-agent detection-revision-critique pipeline. Use when: 'improve text from few examples', 'learn revision patterns from feedback', 'build experience-based text correction', 'distill feedback into reusable rules', 'small-sample text improvement pipeline', 'create agentic revision system from examples'.
    14 repo stars
  90. ▌
    Orthogonal Hierarchical Decomposition Structure Aw · ndpvt-web
    Decompose complex tables with multi-level headers, merged cells, and irregular layouts into orthogonal column/row trees for structure-aware question answering with LLMs. Use when: 'analyze this complex table', 'answer questions about this hierarchical table', 'parse table with merged cells', 'extract data from nested headers', 'table QA with multi-level columns', 'understand this pivot table structure'.
    14 repo stars
  91. ▌
    Pearl Prototype Enhanced Alignment Label Efficient · ndpvt-web
    Implements PEARL (Prototype-Enhanced Aligned Representation Learning) to improve embedding quality for nearest-neighbor retrieval, similarity search, and lightweight classifiers when labeled data is scarce. Reshapes embedding geometry by softly aligning vectors toward class prototypes without retraining the base encoder. Trigger phrases: - "improve my embedding quality for retrieval" - "fix nearest neighbor search accuracy with few labels" - "align embeddings toward class prototypes" - "post-process embeddings for better similarity search" - "label-efficient embedding refinement" - "improve kNN classification with limited labels"
    14 repo stars
  92. ▌
    Ral Bench Benchmarking Application Level Functiona · ndpvt-web
    Generate and evaluate complete multi-file application repositories with both functional correctness and non-functional quality (maintainability, security, robustness, efficiency, resource usage). Use when: 'generate a complete application from requirements', 'build a multi-file project', 'evaluate code quality of a repository', 'check application-level correctness', 'assess non-functional quality attributes', 'scaffold a production-grade project with tests'.
    14 repo stars
  93. ▌
    Reasoning Augmented Representations Multimodal Ret · ndpvt-web
    Decouple reasoning from embedding compression in multimodal retrieval pipelines by enriching queries and corpus entries with explicit semantic context before encoding. Use when: 'build a multimodal search system', 'improve image-text retrieval accuracy', 'fix retrieval for ambiguous queries', 'add reasoning to my embedding pipeline', 'my CLIP search returns wrong results for complex queries', 'enhance retrieval with VLM captions'.
    14 repo stars
  94. ▌
    Swe Agi Benchmarking Specification Driven Software · ndpvt-web
    Build production-scale software systems from formal specifications, RFCs, and standards documents using specification-driven construction methodology. Triggers: 'implement this RFC', 'build a parser from the spec', 'implement this standard', 'construct from specification', 'build from RFC', 'specification-driven implementation'
    14 repo stars
  95. ▌
    Swe Refactor Repository Level Benchmark Real World · ndpvt-web
    Perform repository-level code refactoring with semantics-preserving guarantees using the SWE-Refactor methodology. Supports atomic refactorings (Extract Method, Move Method, Inline Method) and compound refactorings (Extract+Move, Move+Rename, Move+Inline) with cross-file dependency tracking and multi-stage verification. Trigger phrases: "refactor this method across the repo", "extract method from this function", "move this method to another class", "inline this method", "refactor without breaking tests", "restructure this code while preserving behavior"
    14 repo stars
  96. ▌
    Timemachine Bench Benchmark Evaluating Capabilitie · ndpvt-web
    Systematic dependency migration for Python projects. Diagnose and fix test failures caused by dependency updates using a minimal-edit, test-driven agent loop. Triggers: 'migrate dependencies', 'fix breaking changes from package update', 'upgrade library version', 'dependency migration', 'tests broke after pip upgrade', 'adapt code to new API version'
    14 repo stars
  97. ▌
    Variability Aware Detection Repair Compilation Err · ndpvt-web
    Detect and repair compilation errors hidden behind #ifdef/#ifndef/#if defined() preprocessor directives in configurable C/C++ systems. Analyzes all feature combinations to find errors that only manifest under specific configurations. Use when: 'check my C code for ifdef errors', 'find compilation bugs across configurations', 'fix variability-induced compilation errors', 'analyze preprocessor conditionals for hidden bugs', 'audit configurable system for compile failures', 'detect errors in Linux kernel config-dependent code'.
    14 repo stars
  98. ▌
    Vision Deepresearch Incentivizing Deepresearch Cap · ndpvt-web
    Multi-turn, multi-entity, multi-scale visual and textual deep research agent for answering complex questions about images. Implements the Vision-DeepResearch paradigm: iterative reasoning-then-search with progressive visual cropping and text retrieval. Use when: 'research this image', 'identify everything in this photo', 'what is the history behind this building', 'find information about objects in this picture', 'deep research this visual scene', 'multi-hop visual question answering'.
    14 repo stars
  99. ▌
    Vulread Knowledge Graph Guided Software Vulnerabil · ndpvt-web
    CWE-guided vulnerability reasoning and detection using knowledge-graph-structured analysis. Analyzes source code for security vulnerabilities with structured CWE-level explanations grounded in a security knowledge graph. Triggers: 'analyze this code for vulnerabilities', 'find CWE issues in this function', 'security audit with CWE classification', 'explain the vulnerability in this code', 'what CWE does this bug fall under', 'check this C function for memory safety issues'
    14 repo stars
  100. ▌
    Xlist Hate Checklist Based Framework Interpretable · ndpvt-web
    Decompose hate speech detection into a checklist of ten concept-level binary questions answered independently by an LLM, then aggregate results via a lightweight decision tree for interpretable, cross-dataset-robust classification. Use when asked to: 'build a hate speech detector', 'create an interpretable content moderation system', 'detect hateful content with explainability', 'classify toxic text with audit trails', 'implement checklist-based text classification', 'make a robust hate speech pipeline'.
    14 repo stars