Productivity

nota-america Skill

Move a Google Calendar event to a new time and automatically notify all attendees.

nota-america Skill

Google Workflow: Convert a Gmail message into a Google Tasks entry.

nota-america Skill

Google Workflow: Announce a Drive file in a Chat space.

nota-america Skill

Google Workflow: Weekly summary: this week's meetings + unread email count.

nota-america Skill

Read content from a Google Doc and use it as the body of a Gmail message.

nota-america Skill

Review your Google Calendar week, identify gaps, and add events to fill them.

nota-america Skill

Share a Google Docs document with edit access and email collaborators the link.

nota-america Bundle

Create a recurring daily UI inspiration capture. Use when the user asks to run, refresh, package, or validate dated UI inspiration bundles, especially for `articles/YYYY-MM-DD-ui-inspiration-capture/` outputs, Framer/Dribbble landing-page inspiration, motion-study screenshots/videos, AI-builder prompts, duplicate checking, or converting a project runbook into repeatable workflow.

nota-america Skill

Add a list of attendees to an existing Google Calendar event and send notifications.

adu2021 Skill

Tool-R0 trains tool-use agents through self-play between Generator and Solver, creating curriculum-like progression without manual task curation.

adu2021 Skill

Train efficient robot manipulation policies by strategically applying task diversity and debiasing expert demonstrations to remove execution speed variations that degrade learning.

adu2021 Skill

Bridge vision-to-action gap using pointing as unified intermediate representation, enabling 56.2% success on manipulation tasks without task-specific fine-tuning.

adu2021 Skill

Build knowledge agents that generalize across grounded reasoning tasks by combining agentic data synthesis via vector search with off-policy reinforcement learning. Use multi-task training on synthetic question-answer pairs to achieve robust cross-domain performance.

adu2021 Skill

Enable native parallel token generation in language models by implementing adaptive task decomposition and merge strategies, achieving 2x speedup with 1.87% performance gains.

adu2021 Skill

Enable web agents to improve over time by curating episodic memory from navigation trajectories and dynamically injecting task-specific guidance at runtime—no retraining required, persistent improvement across sessions.

adu2021 Skill

Train language models for multi-step information-seeking using dual-constrained task synthesis and cost-efficient staged learning. Generate complex queries by controlling topological complexity and information dispersion, then train atomic reasoning skills before introducing environment interaction. Achieve state-of-the-art on long-horizon search benchmarks with 30B parameter models.

adu2021 Skill

Scale computer-use agents from 30% to 72% success rate by generating parallel rollouts and selecting best trajectories through behavior narrative evaluation. Use when deploying desktop agents on complex, high-variance task scenarios.

adu2021 Skill

Unified MLLM processing four urban data types simultaneously: geospatial structures, trajectory information, satellite imagery, and street-view photos. Outperforms general-purpose models on 12-task urban benchmark with 31-375% improvements. Use for urban planning, traffic analysis, location intelligence, and smart city applications requiring integrated spatial reasoning.

adu2021 Skill

Build scalable open-source agents that autonomously operate graphical user interfaces across Windows, macOS, Linux, Android, iOS, and web platforms using large-scale cross-platform training data and vision-language models for GUI understanding, element grounding, and task completion.

adu2021 Skill

Dynamically allocate reasoning budgets per-task: 36.8% length reduction while improving accuracy 8.4% by learning which problems warrant deep reasoning vs. quick answers. Use when optimizing reasoning efficiency across variable-difficulty problems.

adu2021 Skill

Adaptively prune visual tokens from vision encoders by reconstructing discarded features from retained ones, reducing computational cost by 50% while maintaining task performance on OCR and image understanding tasks.

adu2021 Skill

Inject privileged hints during GRPO training to reshape rollout distributions when advantage collapses occur, increasing outcome diversity without changing task rewards. Hints removed at deployment; policy automatically learns when to use hints via online refresh mechanism.

adu2021 Skill

Enable richer visual reasoning in multimodal models via learnable latent tokens trained with visual bottlenecking. Introduces special tokens that implicitly learn task-relevant visual abstractions without explicit supervision, through attention masking forcing visual information through latents—improving vision-heavy tasks without task-specific annotations.

adu2021 Skill

Combines DeepSeek Sparse Attention (DSA) achieving O(Lk) complexity, scalable RL framework allocating 10%+ compute to post-training, and large-scale agentic task synthesis with 1,800+ environments. Matches proprietary models in reasoning; DeepSeek-V3.2-Speciale achieves olympiad gold.

adu2021 Skill

Optimize data mixtures across diverse vision-language domains when applying RL with verifiable rewards to multimodal LLMs, balancing task-specific performance with generalization.

adu2021 Skill

Enable balanced multi-task GRPO training via robustness-aware optimization and improvement-aware task reweighting, dynamically adjusting task weights based on both reward and loss trajectory improvement, achieving 6-28% worst-task improvements while maintaining competitive average accuracy.

adu2021 Skill

Enables continuous self-improvement through corpus-grounded self-play. Challenger mines difficult examples from document corpus for Reasoner to solve. External corpus prevents task stagnation that plagues closed-loop self-play. Achieves 8.9% math, 9.8% general reasoning improvements.

adu2021 Skill

Scale information-seeking agents via nested architecture separating outer reasoning from inner page exploration. Minimal toolkit (search, visit, click, fill) handles dynamic web interactions, inner loop filters content before context injection, multi-task learning trains outer/inner jointly—enabling 4B models to match larger competitors.

adu2021 Skill

Achieve post-hoc debiasing of frozen vision-language models by operating in sparse autoencoder latent space. Identify and modulate bias-relevant neurons with neuron-level precision while preserving task-relevant features, requiring no task-specific fine-tuning.

adu2021 Skill

Mitigate lazy reasoning in Large Reasoning Models via self-distillation teaching task decomposition, followed by Diversity-Aware GRPO with entropy-based advantage functions, enabling effective decomposition without external teachers while balancing structured reasoning with diversity.

nota-america Bundle

Builds end-to-end hiring packets — job post, structured interview guide with scoring rubric, and offer letter template — from a hiring brief. Triggers on: "help me hire", "we're hiring for", "write a job post", "job description", "JD", "open role", "create a job ad", "interview questions", "scoring rubric", "draft an offer letter", "send an offer", "make a hiring packet", or any request to recruit for a position. When in doubt, trigger — covers the full hiring workflow from job post through DocuSign envelope creation via browser. Does NOT screen or rank applicants.

nota-america Bundle

Keeps HubSpot current without the owner opening it: creates and updates contacts and deals from email and calendar context, logs notes and calls, and flags stale records. The "stop doing data entry" skill. Use when the user asks to update the CRM, log a call, clean up HubSpot, or add context to a deal.

nota-america Skill

Analyze and improve business processes. Trigger with "this process is slow", "how can we improve", "streamline this workflow", "too many steps", "bottleneck", or when the user describes an inefficient process they want to fix.

nota-america Skill

Handles an incoming customer complaint end-to-end — pulls context, drafts a response, and suggests an operational fix. Accepts optional email or ticket ID argument.

undermybelt Bundle

Use when you have a spec or requirements for a multi-step task, before touching code

undermybelt Bundle

Gmail, Calendar, Drive, Docs, Sheets via gws CLI or Python.