[upstream-sync] README.md from ARUNAGIRINATHAN-K/awesome-ai-agents-2026@e9277078 [catalogue]

upstream/ARUNAGIRINATHAN-K-awesome-ai-agents-2026/catalogue/.github/workflows/link-check.yml

@@ -2,9 +2,9 @@
 title: "Link Check"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/ARUNAGIRINATHAN-K/awesome-ai-agents-2026/blob/117f624b/.github/workflows/link-check.yml
-upstream_sha: 117f624b
-imported_at: 2026-07-25
+upstream_source: https://github.com/ARUNAGIRINATHAN-K/awesome-ai-agents-2026/blob/539ec1dc/.github/workflows/link-check.yml
+upstream_sha: 539ec1dc
+imported_at: 2026-06-26
 prompt_class: unknown
 upstream_changes: accepted
 author: upstream
@@ -40,7 +40,7 @@ jobs:
           args: >-
             --verbose
             --no-progress
-            --accept 200,204,301,302,308,403,429
+            --accept 200,204,301,302,403,429
             --exclude-path node_modules
             --exclude "star-history.com"
             --exclude "shields.io"

upstream/ARUNAGIRINATHAN-K-awesome-ai-agents-2026/catalogue/README.md

@@ -2,9 +2,9 @@
 title: "Awesome AI Agents 2026"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/ARUNAGIRINATHAN-K/awesome-ai-agents-2026/blob/af105c56/README.md
-upstream_sha: af105c56
-imported_at: 2026-08-10
+upstream_source: https://github.com/ARUNAGIRINATHAN-K/awesome-ai-agents-2026/blob/e9277078/README.md
+upstream_sha: e9277078
+imported_at: 2026-07-24
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -127,7 +127,6 @@ validated: false
 - [opencode](https://github.com/anomalyco/opencode) `🌱` `[TypeScript]` `[Desktop]` - Open-source coding agent available as a desktop app with a visual interface.
 - [OpenHands](https://github.com/OpenHands/OpenHands) `🌱` `[Python]` `[Docker]` - AI-driven development platform that writes, tests, and deploys code autonomously.
 - [PR-Agent](https://github.com/The-PR-Agent/pr-agent) `🚀` `[Python]` `[GitHub]` - Open-source AI PR reviewer that auto-describes, reviews, and improves pull requests.
-- [Prime Agent](https://github.com/PrimeIntellect-ai/prime-agent) `🚀` `[TypeScript]` `[CLI]` - Open-source RLM coding and research agent designed for long-running autonomous tasks.
 - [Qodo](https://www.qodo.ai) `🚀` `[Cloud]` `[Security]` - AI code review platform with context-aware PR validation and security analysis.
 - [RooCode](https://github.com/RooCodeInc/Roo-Code) `🌱` `[TypeScript]` `[VS Code]` - Cline fork with structured modes and reduced hallucinations for more reliable code generation.
 - [Snyk Code](https://snyk.io/product/snyk-code/) `🌱` `[Cloud]` `[Security]` - AI-powered security scanner with real-time vulnerability detection in agent-generated code.
@@ -235,9 +234,9 @@ Sandboxes, web scrapers, browser automation, and networking layers that agents d
 
 - [Agent Bounties](https://github.com/NSPG13/agent-bounties) `🔬` `[Rust]` `[MCP]` - Coordinates verifiable digital bounty workflows designed for agents to post, fund, claim, solve, verify, and earn.
 - [AgentDock](https://github.com/agentdock/agentdock) `🚀` `[Python]` `[Docker]` - Framework for building and deploying production-ready AI agents with composable node architecture.
+- [AgentServices](https://agentservices.to) `🚀` `[Python]` `[x402]` - Paid data APIs for AI agents with 54 services, 37 MCP tools, and x402 nanopayments on Base. Market data, onchain analytics, AI inference, and research.
+- [AgentWork](https://agentwork-api.yfoob.chatgpt.site/) `🔬` `[Cloud]` `[x402]` - Aggregates verified paid work opportunities for autonomous agents, with a 0.005 Polygon USDC/hour x402 API for full decision context.
 - [codex-profiles](https://github.com/Ducksss/codex-profiles) `🚀` `[Python]` `[OpenAI]` - Bash CLI for switching OpenAI Codex CLI and Desktop profiles with isolated CODEX_HOME directories.
-- [Agent Island](https://github.com/tristan666666/agent-island) `🚀` `[Swift]` `[Observability]` - Menu bar and notch companion showing live session state and your-turn alerts for Claude Code, Codex, Gemini, Grok and Cursor, with quota and cost computed locally.
-- [CompozyOS](https://github.com/compozy/compozy) `🚀` `[Go]` `[Multi-Agent]` - Runs agent CLIs as a team on loops and schedules, with shared memory, permissions and approvals in one self-hosted daemon.
 - [Crawl4AI](https://github.com/unclecode/crawl4ai) `🌱` `[Python]` `[Multi-Agent]` - Extracts structured data from web pages using LLM-friendly output formats optimized for agent ingestion.
 - [Docling](https://github.com/docling-project/docling) `🌱` `[Python]` `[IDE]` - Parses PDFs, DOCX, and slides into structured text with deep layout understanding for document agents.
 - [E2B](https://github.com/e2b-dev/e2b) `🌱` `[TypeScript]` `[Multi-Agent]` - Cloud sandboxes for AI agents to run code securely in isolated environments.
@@ -264,8 +263,8 @@ Sandboxes, web scrapers, browser automation, and networking layers that agents d
 - [Flowise](https://github.com/FlowiseAI/Flowise) `🌱` `[TypeScript]` `[RAG]` - Open-source drag-and-drop UI to build customized LLM flows, RAG pipelines, and agent systems.
 - [Langflow](https://github.com/langflow-ai/langflow) `🌱` `[Python]` `[RAG]` - Visual drag-and-drop builder for LLM workflows, RAG agents, and multi-step pipelines.
 - [Lindy](https://www.lindy.ai) `🌱` `[Cloud]` `[No-Code]` - No-code AI agent platform with 3000+ app integrations for business workflow automation.
-- [LlamaIndex Workflows](https://developers.llamaindex.ai/python/llamaagents/workflows/) `🌱` `[Python]` `[RAG]` - Event-driven orchestration framework for building complex agentic systems.
-- [LLMGraph](https://llmgraph.ai/) `🔬` `[Cloud]` `[RAG]` - Visual canvas for building RAG chatbots and AI agents with one-click deploy to REST API or chat widget.
+- [LlamaIndex Workflows](https://docs.llamaindex.ai/en/stable/module_guides/workflow/) `🌱` `[Python]` `[RAG]` - Event-driven orchestration framework for building complex agentic systems.
+- [LLMGraph](https://llmgraph.ai/home) `🔬` `[Cloud]` `[RAG]` - Visual canvas for building RAG chatbots and AI agents with one-click deploy to REST API or chat widget.
 - [Make](https://www.make.com/en) `🌱` `[Cloud]` `[RAG]` - Visual workflow automation platform with AI capabilities and drag-and-drop scenario builder.
 - [n8n](https://github.com/n8n-io/n8n) `🌱` `[TypeScript]` `[Docker]` - Open-source workflow automation with AI agent nodes combining visual and code-based orchestration.
 - [Relevance AI](https://relevanceai.com) `🌱` `[Cloud]` `[No-Code]` - No-code AI agent builder for sales, support, and research use cases with team collaboration.
@@ -280,9 +279,7 @@ Sandboxes, web scrapers, browser automation, and networking layers that agents d
 - [LiveKit Agents](https://github.com/livekit/agents) `🌱` `[Python]` `[IDE]` - Framework for building real-time, multimodal AI agents with voice, video, and data channels.
 - [Nuance AI](https://dragon.nuance.com/en-us/home) `🚀` `[Cloud]` `[CLI]` - Enterprise speech and conversational AI platform for clinical and contact-center workflows with HIPAA-capable deployments.
 - [Google Cloud Speech-to-Text v2](https://cloud.google.com/speech-to-text) `🚀` `[Cloud]` `[Pipeline]` - Google Cloud streaming and batch speech recognition API v2 with improved accuracy, streaming, and noise suppression for real-time agent pipelines.
-- [Personal Jarvis](https://github.com/PersonalJarvis/PersonalJarvis) `🔬` `[Python]` `[Voice]` - Voice-driven desktop assistant that takes mouse and keyboard and delegates heavy tasks to agent harnesses like Claude Code, Codex, and MCP.
 - [Pipecat](https://github.com/pipecat-ai/pipecat) `🚀` `[Python]` `[Streaming]` - Production-grade voice AI framework with sub-250ms latency, WebRTC support, multimodal (voice+vision+text), real-time streaming, and 70+ language support.
-- [qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) `🌱` `[Desktop]` `[Voice]` - Full-duplex voice runtime that drives coding agents (Claude Code, Codex, OpenCode, Kimi Code, and more) over ACP, keeping conversations going while background tasks run, with barge-in and a local wake word.
 - [Rasa](https://github.com/RasaHQ/rasa) `🌱` `[Python]` `[Self-Hosted]` - Open-source conversational AI framework with self-hosted NLU training and dialogue management.
 - [simulate-sdk](https://github.com/future-agi/simulate-sdk) `🌱` `[Python]` `[Voice]` - Persona- and scenario-driven SDK for simulating voice and text AI agents.
 - [Vapi](https://github.com/VapiAI/server-sdk-python) `🌱` `[Python]` `[Voice]` - Platform for building voice AI agents with low-latency speech-to-speech capabilities.
@@ -319,6 +316,7 @@ Sandboxes, web scrapers, browser automation, and networking layers that agents d
 - [Agent Learning Kit](https://github.com/future-agi/agent-learning-kit) `🌱` `[Python]` `[Evaluation]` - LLM evaluation framework with 50+ metrics, LLM-as-Judge, and guardrail scanners (jailbreak, PII, injection).
 - [Shipmoor](https://shipmoor.dev) `🔬` `[Python]` `[Testing]` - Local, deterministic verification layer for AI agent code: scans, test evidence, and a binding merge verdict without uploading source.
 - [SourceryKit](https://github.com/ProvablyAI/sourcerykit) `🔬` `[Python]` `[Security]` - Verifies an agent's outbound requests and MCP handoffs against a source of truth using zero-knowledge proofs, logging each call and blocking anything off the trusted-endpoint allow-list.
+- [ai-evaluation](https://github.com/future-agi/ai-evaluation) `🌱` `[Python]` `[Evaluation]` - LLM evaluation framework with 50+ metrics, LLM-as-Judge, and guardrail scanners (jailbreak, PII, injection).
 - [Future AGI](https://github.com/future-agi/future-agi) `🌱` `[Python]` `[Self-Hosted]` - Self-hostable end-to-end agent engineering platform with tracing, evals, guardrails, and gateway.
 
 ## Agent Interfaces and UIs
@@ -353,7 +351,6 @@ Frontend workspaces and chat interfaces with built-in agent plugins and tool-use
 
 - [AgentBench](https://github.com/THUDM/AgentBench) `🌱` `[Python]` `[Benchmark]` - Comprehensive benchmark for evaluating LLMs as agents across 8 distinct environments.
 - [ARC-AGI-2](https://arcprize.org) `🌱` `[Python]` `[Benchmark]` - Frontier benchmark for measuring general intelligence capabilities in AI agents beyond pattern matching.
-- [ClawBench](https://github.com/TIGER-AI-Lab/ClawBench) `🔬` `[Python]` `[Benchmark]` - Evaluates web agents on 283 real-world tasks across 163 live websites with interception and trace-based scoring.
 - [GAIA Benchmark](https://huggingface.co/papers/2311.12983) `🌱` `[Python]` `[Benchmark]` - Benchmark for General AI Assistants measuring real-world reasoning and tool use.
 - [Inspect AI](https://github.com/UKGovernmentBEIS/inspect_ai) `🌱` `[Python]` `[Evaluation]` - Framework for evaluating large language models with composable tasks and scoring.
 - [SWE-bench](https://github.com/SWE-bench/SWE-bench) `🚀` `[Python]` `[GitHub]` - Benchmark for evaluating LLMs on real-world software engineering tasks from GitHub issues.
@@ -386,8 +383,6 @@ Curated list of vertical agent solutions for finance, healthcare, legal, manufac
 - [JPMorgan AlphaBlue](https://www.jpmorgan.com/global) `🚀` `[Cloud]` `[Multi-Agent]` - Trading and market analysis agent integrations for institutional workflows.
 - [Goldman Sachs Marquee AI](https://www.goldmansachs.com) `🚀` `[Cloud]` `[Multi-Agent]` - Market data and analytics agents built on Marquee platform.
 - [Morgan Stanley AdvisorBot](https://www.morganstanley.com) `🚀` `[Cloud]` `[CLI]` - Financial advisory assistant for advisors and retail clients.
-- [AgentPump](https://agentpump.app) `🔬` `[Cloud]` `[CLI]` - Runs autonomous on-chain memecoin trading agents on Solana that trade on a schedule, operable from the terminal via the @agentpump/cli.
-
 
 ### Healthcare
 
@@ -409,7 +404,7 @@ Curated list of vertical agent solutions for finance, healthcare, legal, manufac
 ### Government & Compliance
 
 - Anthropic Government Agents - Policy analysis and public sector agents for regulated workflows (🏷️ `Cloud` `Government` `Enterprise`).
-- [Leyna](https://www.atom.com/name/Leyna.ai) `🌱` `[Cloud]` `[Government]` - Public records request agent for government agencies to parse, redact, and track requests.
+- [Leyna](https://www.leyna.ai) `🌱` `[Cloud]` `[Government]` - Public records request agent for government agencies to parse, redact, and track requests.
 - Tax & Insurance Agent Platforms - Generic category placeholder for compliance-focused tax and underwriting agents (🏷️ `Cloud` `Compliance` `Enterprise`).
 
 ## Learning Resources
@@ -540,7 +535,7 @@ Tools for generating images, video, music, audio, and 3D assets using AI models.
 - [LTX Video](https://github.com/Lightricks/LTX-Video) `🌱` `[Python]` `[IDE]` - Commercially safe open-source video generation model trained on fully licensed data.
 - [Luma Dream Machine](https://lumalabs.ai/app) `🌱` `[Cloud]` `[IDE]` - Creates physics-accurate 4K HDR video scenes with 3D-style output for cinematic quality.
 - [Pika 2.5](https://pika.art/login) `🌱` `[Cloud]` `[IDE]` - Beginner-friendly video generation tool with fast turnaround and intuitive editing controls.
-- [Runway Gen-4.5](https://runway.com) `🌱` `[Cloud]` `[IDE]` - Professional video generation with Motion Brush and the highest benchmark scores for visual quality.
+- [Runway Gen-4.5](https://runwayml.com) `🌱` `[Cloud]` `[IDE]` - Professional video generation with Motion Brush and the highest benchmark scores for visual quality.
 - [Seedance 2.0](https://seed.bytedance.com/en/seedance) `🌱` `[Cloud]` `[IDE]` - Quad-modal input video generator with built-in lip-sync and 2K resolution output.
 - [Sora 2](https://sora.com) `🚀` `[Cloud]` `[OpenAI]` - Generates 25-second cinematic video clips with strong narrative coherence from text prompts by OpenAI.
 - [Wan 2.1](https://github.com/Wan-Video/Wan2.1) `🌱` `[Python]` `[IDE]` - Open-source self-hosted video generation model with unlimited length and no per-video costs.
@@ -574,9 +569,9 @@ AI agents that automate customer support, CRM workflows, sales outreach, and tic
 
 - [Ada](https://www.ada.cx) `🚀` `[Cloud]` `[Multi-Agent]` - Resolves 60% of customer support tickets automatically with complex multi-turn query understanding.
 - [Assembled](https://www.assembled.com) `🚀` `[Cloud]` `[Multi-Agent]` - Routes support tickets with workforce-aware scheduling and intelligent handoff to human agents.
-- [ChatBotKit](https://chatbotkit.com) `🌱` `[Cloud]` `[RAG]` - Deploys AI agents that answer customer questions from business data and hand conversations to humans.
 - [Dixa](https://www.dixa.com) `🚀` `[Cloud]` `[Multi-Agent]` - CRM-first conversational support platform with AI-powered routing and customer context enrichment.
 - [Freshdesk Freddy AI](https://www.freshworks.com/freshdesk/omni/freddy-ai-automation/) `🌱` `[Cloud]` `[Multi-Agent]` - Auto-triages and routes support tickets with smart AI suggestions for budget-conscious SMB teams.
+- [Hellomatik](https://hellomatik.com) `🌱` `[Cloud]` `[Multi-Agent]` - Turns company knowledge into agents that answer, book, and sell across WhatsApp, phone, email, and web.
 - [Intercom Fin](https://fin.ai) `🚀` `[Cloud]` `[Multi-Agent]` - Resolves 50% of SaaS support tickets by learning directly from your help center and knowledge base.
 - [Zendesk AI](https://www.zendesk.com/service/ai/) `🚀` `[Cloud]` `[Multi-Agent]` - Automates 30% of enterprise support tickets with deep integration into the existing Zendesk ecosystem.
 
@@ -600,7 +595,6 @@ AI agents that automate customer support, CRM workflows, sales outreach, and tic
 - [Clay](https://www.clay.com) `🌱` `[Cloud]` `[IDE]` - Enriches leads from 70+ data providers and generates hyper-personalized outreach at scale.
 - [Instantly](https://instantly.ai) `🌱` `[Cloud]` `[Multi-Agent]` - Generates AI cold emails with smart sender rotation and built-in domain warmup for deliverability.
 - [Lavender](https://lavender.ai/) `🌱` `[Cloud]` `[Multi-Agent]` - Coaches email writing in real-time with AI response scoring and recipient intelligence.
-- [OutreachAgent](https://outreachagent.dev/for-agents) `🔬` `[Cloud]` `[Event-Driven]` - Runs reply-aware outbound email workflows with webhooks, sender pacing, approvals, and deliverability guardrails.
 - [Overloop CLI](https://overloop.com) `🌱` `[Cloud]` `[CLI]` - AI outbound CLI agent that sources 450M+ contacts and runs email plus LinkedIn campaigns with JSON output.
 
 ## Voice Agent Platforms
@@ -625,7 +619,6 @@ Platforms for building, deploying, and scaling voice-based AI agents across call
 - [Synthesia](https://www.synthesia.io) `🌱` `[Cloud]` `[IDE]` - Generates AI video avatars that speak in 120+ languages for training and communication agents.
 - [Synthflow](https://synthflow.ai) `🌱` `[Cloud]` `[No-Code]` - No-code voice agent builder with pre-built templates for SMBs to deploy phone agents quickly.
 - [Voiceflow](https://www.voiceflow.com) `🌱` `[Cloud]` `[No-Code]` - No-code builder for voice and chat agents with visual conversation design and team collaboration.
-- [Workforce Wave](https://www.workforcewave.com/) `🔬` `[Cloud]` `[Voice]` - AI voice receptionist for SMBs handling 24/7 call answering, appointment booking, and lead capture.
 
 ## Deep Research Agents
 
@@ -837,6 +830,7 @@ Curated newsletters, podcasts, and communities for staying current with AI agent
 - [The Agents Index](https://theagentsindex.com) `🔬` `[Cloud]` `[Multi-Agent]` - Compares AI agent tools with sourced pricing, verdicts, and pros/cons in a researched, quality-gated directory.
 - [The Rundown AI](https://www.therundown.ai) `🌱` `[Python]` `[RAG]` - Daily AI digest reaching 600K+ subscribers with concise coverage of agent news and launches.
 
+- [Agents Launchpad](https://launchpad.smartbizcalc.com) `🌱` `[Python]` `[Multi-Agent]` - Community-curated directory of indie AI agents and tools.
 ## Changelog
 
 See [CHANGELOG.md](CHANGELOG.md) for the full update history.

upstream/VoltAgent-awesome-ai-agent-papers/catalogue/README.md

@@ -2,9 +2,9 @@
 title: "Awesome AI Agent Papers"
 task: ""
 lineage_type: import
-upstream_source: https://github.com/VoltAgent/awesome-ai-agent-papers/blob/c8502b6a/README.md
-upstream_sha: c8502b6a
-imported_at: 2026-08-08
+upstream_source: https://github.com/VoltAgent/awesome-ai-agent-papers/blob/aa50c0c3/README.md
+upstream_sha: aa50c0c3
+imported_at: 2026-07-03
 prompt_class: catalogue
 upstream_changes: accepted
 author: upstream
@@ -69,13 +69,12 @@ Hundreds of papers are published on arXiv every week, and a growing number of th
 <br>
 
 <details open id="multi-agent">
-<summary><h3 style="display:inline">Multi-Agent (54)</h3></summary>
+<summary><h3 style="display:inline">Multi-Agent (53)</h3></summary>
 
 <br>
 
 | Paper | arXiv ID |
 |---|:---:|
-| **[Agon: An Autonomous Large-Scale Omnidisciplinary Research System Built on Prompt Economy](https://arxiv.org/pdf/2606.24177)** - Treats prompt engineering as an engineering discipline rather than an art, minimizing human time while expecting maximum output. Carries the smallest prompt footprint among cross-disciplinary autoresearch systems (18 roles, 230.6 KiB total), has run across 10+ research fields without modification, and the longest observed run went 30 days unattended. | <a href="https://arxiv.org/abs/2606.24177"><img src="https://img.shields.io/badge/arXiv-2606.24177-b31b1b.svg" alt="arXiv" /></a> |
 | **[PerceptUI: LLM Agents as Human-Aligned Synthetic Users for UI/UX Evaluation](https://arxiv.org/pdf/2606.05697)** - A persona-conditioned framework that predicts how a specific user would answer UI/UX evaluation questions and explains why in natural language. Trained via contrastive reflection fine-tuning and reflective prompt evolution, reaching human-level realism and generalizing to unseen questions and personas. | <a href="https://arxiv.org/abs/2606.05697"><img src="https://img.shields.io/badge/arXiv-2606.05697-b31b1b.svg" alt="arXiv" /></a> |
 | **[AutoNumerics: An Autonomous, PDE-Agnostic Multi-Agent Pipeline for Scientific Computing](https://arxiv.org/pdf/2602.17607)** - A multi-agent pipeline that reads a PDE problem description in plain text and writes, debugs, and validates a classical numerical solver end-to-end. Generates spectral and finite-difference code (no neural networks), scoring ~6 orders of magnitude below FNO and CodePDE baselines. | <a href="https://arxiv.org/abs/2602.17607"><img src="https://img.shields.io/badge/arXiv-2602.17607-b31b1b.svg" alt="arXiv" /></a> |
 | **[Beyond Offline A/B Testing: Context-Aware Agent Simulation for Recommender System Evaluation](https://arxiv.org/abs/2604.09549)** - Evaluates recommender systems via agent-RS interactions. | <a href="https://arxiv.org/abs/2604.09549"><img src="https://img.shields.io/badge/arXiv-2602.06039-b31b1b.svg" alt="arXiv" /></a> |
@@ -207,7 +206,7 @@ Hundreds of papers are published on arXiv every week, and a growing number of th
 <br>
 
 <details id="eval--observability">
-<summary><h3 style="display:inline">Eval & Observability (81)</h3></summary>
+<summary><h3 style="display:inline">Eval & Observability (80)</h3></summary>
 
 <br>
 
@@ -215,7 +214,6 @@ Hundreds of papers are published on arXiv every week, and a growing number of th
 |---|:---:|
 | **[PerspectiveGap: A Benchmark for Multi-Agent Orchestration Prompting](https://arxiv.org/pdf/2606.08878)** - A 110-scenario benchmark testing whether LLMs can compose orchestration prompts that distribute context to sub-agents without information leakage. Across 10 topologies and 27 commercial models, GPT-5.5 leads with 62% pass rate while the average is 14.9%. | <a href="https://arxiv.org/abs/2606.08878"><img src="https://img.shields.io/badge/arXiv-2606.08878-b31b1b.svg" alt="arXiv" /></a> |
 | **[ClawBench: Evaluating Browser Agents on Live Production Websites with Submission-Interception](https://arxiv.org/abs/2604.08523)** - Benchmarks browser agents on 283 everyday tasks (V1 153 + V2 130) across 163 live production sites, with a Chrome-extension plus CDP layer that blocks only the final write request so agents can run end-to-end on real sites without real-world side effects. Two-stage scoring (interception + LLM judge); leaderboard at https://claw-bench.com. | <a href="https://arxiv.org/abs/2604.08523"><img src="https://img.shields.io/badge/arXiv-2604.08523-b31b1b.svg" alt="arXiv" /></a> |
-| **[StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs](https://arxiv.org/pdf/2505.20139)** - Benchmarks structured-output generation and cross-format conversion across 18 text and renderable formats, with syntax, structural, and visual evaluation checks. | <a href="https://arxiv.org/abs/2505.20139"><img src="https://img.shields.io/badge/arXiv-2505.20139-b31b1b.svg" alt="arXiv" /></a> |
 | **[From Features to Actions: Explainability in Traditional and Agentic AI Systems](https://arxiv.org/pdf/2602.06841v1)** - Compares attribution-based explanations with trace-based diagnostics across static and agentic settings to study how explainability methods translate to multi-step agent trajectories. | <a href="https://arxiv.org/abs/2602.06841v1"><img src="https://img.shields.io/badge/arXiv-2602.06841-b31b1b.svg" alt="arXiv" /></a> |
 | **[Agentic Uncertainty Reveals Agentic Overconfidence](https://arxiv.org/pdf/2602.06948v1)** - Investigates whether agents can accurately predict their own success rates in agentic tasks. | <a href="https://arxiv.org/abs/2602.06948v1"><img src="https://img.shields.io/badge/arXiv-2602.06948-b31b1b.svg" alt="arXiv" /></a> |
 | **[AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents](https://arxiv.org/pdf/2602.06855v1)** - Introduces 20 research tasks from real ML papers covering idea generation, experiments, and refinement for benchmarking science agents. | <a href="https://arxiv.org/abs/2602.06855v1"><img src="https://img.shields.io/badge/arXiv-2602.06855-b31b1b.svg" alt="arXiv" /></a> |
@@ -260,7 +258,6 @@ Hundreds of papers are published on arXiv every week, and a growing number of th
 | **[The Why Behind the Action: Unveiling Internal Drivers via Agentic Attribution](https://arxiv.org/pdf/2601.15075v2)** - Proposes a hierarchical framework for general agentic attribution that identifies internal factors driving agent actions through temporal likelihood dynamics and perturbation-based analysis. | <a href="http://arxiv.org/abs/2601.15075v2"><img src="https://img.shields.io/badge/arXiv-2601.15075-b31b1b.svg" alt="arXiv" /></a> |
 | **[Tokenomics: Quantifying Where Tokens Are Used in Agentic Software Engineering](https://arxiv.org/pdf/2601.14470v1)** - Analyzes token consumption patterns across software development lifecycle stages in a multi-agent system to identify where tokens are consumed and which stages drive cost. | <a href="http://arxiv.org/abs/2601.14470v1"><img src="https://img.shields.io/badge/arXiv-2601.14470-b31b1b.svg" alt="arXiv" /></a> |
 | **[APEX-Agents](https://arxiv.org/pdf/2601.14242v2)** - Introduces a benchmark of 480 long-horizon, cross-application productivity tasks created by investment banking analysts, consultants, and lawyers for evaluating AI agent capabilities in realistic work environments. | <a href="http://arxiv.org/abs/2601.14242v2"><img src="https://img.shields.io/badge/arXiv-2601.14242-b31b1b.svg" alt="arXiv" /></a> |
-| **[DRIFT: Detecting Representational Inconsistencies for Factual Truthfulness](https://arxiv.org/pdf/2601.14210)** - Lightweight probes (3M-37M params) trained on middle-layer hidden states catch factually wrong generations before they reach the user, at less than 0.1% overhead. Up to 13 AUROC points above final-layer baselines, and the probes transfer across datasets without retraining. | <a href="https://arxiv.org/abs/2601.14210"><img src="https://img.shields.io/badge/arXiv-2601.14210-b31b1b.svg" alt="arXiv" /></a> |
 | **[CooperBench: Why Coding Agents Cannot be Your Teammates Yet](https://arxiv.org/pdf/2601.13295v2)** - Introduces a benchmark of 600+ collaborative coding tasks to evaluate whether coding agents can coordinate as effective teammates under various coordination structures. | <a href="http://arxiv.org/abs/2601.13295v2"><img src="https://img.shields.io/badge/arXiv-2601.13295-b31b1b.svg" alt="arXiv" /></a> |
 | **[Insider Knowledge: How Much Can RAG Systems Gain from Evaluation Secrets?](https://arxiv.org/pdf/2601.13227v1)** - Investigates how RAG systems can game nugget-based LLM judge evaluations through metric overfitting, demonstrating near-perfect scores when evaluation elements are leaked or predictable. | <a href="http://arxiv.org/abs/2601.13227v1"><img src="https://img.shields.io/badge/arXiv-2601.13227-b31b1b.svg" alt="arXiv" /></a> |
 | **[Replayable Financial Agents: A Determinism-Faithfulness Assurance Harness for Tool-Using LLM Agents](https://arxiv.org/pdf/2601.15322v1)** - Introduces the Determinism-Faithfulness Assurance Harness for measuring trajectory determinism and evidence-conditioned faithfulness in tool-using LLM agents across 74 configurations and 12 models. | <a href="http://arxiv.org/abs/2601.15322v1"><img src="https://img.shields.io/badge/arXiv-2601.15322-b31b1b.svg" alt="arXiv" /></a> |