Machine learning

Point at performance-defect and memory-leak candidates in a diff — a read-only LLM reviewer reads the change plus the repository around it and reports…

Multi-step browser automation via browser-use — the open-source framework for autonomous browser navigation (91K+ stars, MIT).

Use only when creating new registrable ML components that require Factory or Registry patterns.

Retrieve relevant excerpts from a project's text files without loading entire files into context.

Govern Project Jennifer work through stateless-renter routing, situational CDP/CCP transitions, CEEP, POC-vs-FOC, VOC registry parsing, runtime mutation…

将学术论文PDF转化为结构化、可视化的精讲文档。 触发场景: (1) 用户提供PDF论文文件并要求讲解/分析 (2) 用户询问"帮我讲解这篇论文"、"分析这个PDF" (3) 用户需要提取论文的核心方法、实验结果 (4) 用户希望生成论文的可视化HTML摘要 支持…

Automated build orchestration system.

Build, extend, AND query a persistent LLM-maintained wiki for any research topic.

Deep CTF technique atlas for mapping challenge clues to concrete attack patterns, tools, scripts, and next probes across Web, Crypto, Forensics, Reverse, Pwn…

Perform CWE-grounded security code reviews and precise weakness mapping using a locally derived MITRE CWE corpus, relationship graphs, mapping notes…

Use LM Studio as a local LLM backend with Pydantic AI for structured output extraction.

Developer reference for connecting to OpenAI, Anthropic, Gemini, DeepSeek, GLM, MiniMax, MiMo, Mistral, xAI, Qwen, Kimi, Cohere, OpenRouter, Together, and…

Turns a compute run's energy evidence into a typed receipt (MEASURED from counter samples, REPORTED from vendor totals or TDP estimates, UNAVAILABLE when…

Scores binary predictions (Brier, log loss, ECE, MCE, AUROC, accuracy, reliability bins, per-cohort) or saved categorical outputs joined to retained targets…

Builds generative AI applications on Amazon Bedrock.

Rendered guardrail framework: keep a repo's hard rules in one-rule-per-file GRL sources behind a compact router, and render only the rules a task needs.

用于评估视频生成任务的最终产物 (final.mp4) 是否符合 rubric.json 中 output_rubric 列出的指标 (O2-O8 需求达成度/参考忠实度/视觉一致性/音频一致性/物理异常/剧情连贯/音画同步)。当用户要求对一个 case 目录、产物目录、生成结果做 output rubric…

Browser automation for AI agents via inference.sh.

Detects anomalous authentication patterns using UEBA analytics, statistical baselines, and machine learning models to identify impossible travel, credential…

Evaluate an agent skill against a baseline (v3) - trigger tests (explicit, implicit, guidance, negative controls), runs with and without the skill or old vs…

Transcribe audio files or YouTube videos via subtitles or OpenRouter's multimodal API, generate 100-word and 400-word summaries using cloud models…

構成図・アーキテクチャ図・エージェント構成・処理の流れを「動く画像」にして、X（Twitter）などに貼れる MP4 と GIF で書き出すスキル。ノードが順に現れ、線が引かれ、光の粒がエッジを流れ、いま動いている部分だけが光り、下に字幕がタイプされ、最後に全体が同時に動いてループする。箇条書きのツリー（「>…

Chess improvement coach for the operator's 300→1500 chess.com goal.

Smart Copilot PR-review watchdog (multi-repo) that reads review state + timestamps, checks merge conflicts first, and lets an LLM decide…