AI & ML

iblai Skill

Manage an ibl.ai agent's prompts via the platform API — built-in system/proactive/study/guided prompts and greeting settings (saved through the agent settings endpoint), plus full CRUD on suggested prompts. Use when editing what an agent says and suggests.

iblai Skill

Configure an ibl.ai agent's moderation and safety systems via the platform API — enable flags, system prompts and responses (saved through the agent settings endpoint) — and review or delete flagged-prompt moderation logs. Use when setting guardrails or auditing flagged content.

iblai Skill

Manage an ibl.ai agent's training datasets (RAG) via the platform API — list training docs, add resources (file, URL, YouTube, Blackboard, website crawl, GitHub), train/untrain, set visibility and retrain schedule, and delete. Use when feeding an agent knowledge.

iblai Skill

Read an ibl.ai agent's conversation history and summaries via the platform API — list conversations with sentiment/topic/user/date filters, get general summaries and per-conversation memory, and export chat history as an async report. Use when reviewing or exporting how users have chatted with an agent.

iblai Skill

Configure an ibl.ai agent's Privacy Router via the platform API — enable PII detection, choose redact/mask/block action, select entity types (PERSON, EMAIL_ADDRESS, US_SSN, …), set the privacy response, and toggle output filtering (saved through the agent settings endpoint). Use when controlling how the agent handles personal data.

iblai Bundle

Connect/disconnect an ibl.ai agent's sandbox (Claw) instance, manage Claw instances, push config, run health/connectivity checks, and set the agent's model — via the platform API. Use when wiring an agent to a sandboxed workspace.

iblai Bundle

Talk to a deployed ibl.ai agent directly over REST/SSE (or WebSocket) and manage its chat sessions — POST a prompt to the agent chat endpoint, attach arbitrary metadata (surfaced later as client_context), and list/read sessions and per-task history exports. The direct-transport counterpart to iblai-api-agent-chat's MCP wiring; use when you want raw streamed chat + session records rather than an MCP server.

iblai Skill

Read and write an ibl.ai agent's core settings via the platform API — name, description, category, profile image, visibility, and feature flags (anonymous, featured, LTI, attachments, voice, memory, multi-query RAG, forkable). Also fork/duplicate and delete an agent. Use when configuring an agent's basic identity and capabilities.

iblai Skill

Manage an ibl.ai agent's human-support tickets via the platform API — list and filter the tickets users raised with an agent (by agent, requester, status, session), read a ticket's conversation thread, reply as the support team, change ticket status, and close or delete tickets. Use when triaging or responding to support requests escalated from agent chats.

iblai Skill

Manage an ibl.ai agent's disclaimers via the platform API — the Advisory text (saved through the agent settings endpoint) and the User Agreement (create/update/activate). Use when adding legal or advisory notices users must see or accept.

yenhao-huang Bundle

Use for development, debugging, refactoring, testing, benchmarks, and project setup. Follow issue selection or creation, an isolated worktree on an issue-specific agent/fix branch, plan, dev, test, and PR delivery. Use OpenSpec skills for requirements, implementation tasks, tests, and file-change records while preserving project conventions.

yenhao-huang Bundle

Initialize or repair OpenAI Codex MCP configuration for Jina MCP, Firecrawl MCP, Hugging Face MCP, and Git MCP. Use when the user asks to set up MCP servers, add Jina/Firecrawl/Hugging Face/Git MCP to Codex, bootstrap `~/.codex/config.toml`, refresh MCP server blocks, or create a repeatable local MCP initialization workflow.

yenhao-huang Bundle

Run a 4-phase agent loop for optimization tasks: plan, dev, exp, and reflect. Creates an independent timestamped workspace under exp/agent_loop/claude. Each phase is a separate Agent session with fresh context. Parallel subtasks within a phase are dispatched as concurrent subagents. Use when the user says "run agent loop on X", "optimize X with agent loop", "agent loop run X", or resumes a stalled loop with "resume loop" plus a workspace path.

yenhao-huang Skill

Produce fixed-format Traditional Chinese markdown analysis for ESG contest agent-loop runs with plan/dev/exp/reflect records. Use for Claude, generic agent-loop, or non-Codex controllers such as exp/claude_agent_loop and exp/agent_loop_st3_0612.

yenhao-huang Bundle

Create, read, search, update, comment on, label, assign, close, and link GitHub issues. Use when the user asks to file a bug or feature request, manage an existing issue, inspect issue status, work with sub-issues or dependencies, or connect implementation work to an issue. Prefer GitHub MCP tools for reads and use authenticated gh CLI or gh api for writes.

yenhao-huang Bundle

Discover, compare, run, publish, and troubleshoot Hugging Face models, datasets, and Spaces with compatibility, authentication, reproducibility, and safety checks. Use for Hub searches, local inference, gated assets, uploads, fine-tuning preparation, or Hugging Face Python and CLI errors.

yenhao-huang Bundle

Select and verify MCP tools for Swift, SwiftUI, Xcode, profiling, previews, and Apple documentation, and apply the personal apple_query_any_img photo-search project context. Use for iOS tool selection, SwiftUI development workflows, or the ColNomic and Qdrant iPhone photo-search project.

yenhao-huang Bundle

Read, append, update, or restructure Notion pages from local source files through the connected Notion integration, with permission checks and post-write verification. Use for personal Notion publishing and synchronization workflows that must preserve existing content and linked subpages.

yenhao-huang Bundle

Run reproducible local or API-based AI model evaluations across text, architecture, vision, and llama-server speed tasks, preserving per-item responses and source-to-leaderboard traceability. Use when adding a model, rerunning benchmarks, comparing models, diagnosing suspicious scores, or publishing model-evaluation results.

yenhao-huang Skill

Start, debug, and verify local vLLM embedding servers on this machine, especially Qwen embedding models under `/mnt/share_data_78` with the vLLM virtualenv at `/tmp2/howard/venv_manager/vllm`. Use when the user asks to open/start/serve an embedding model with vLLM, mentions qwen-embedding, Qwen3 embedding, first GPU/card, or needs an OpenAI-compatible `/v1/embeddings` endpoint.

yenhao-huang Bundle

Discover and recommend installable agent skills from managed skill libraries. Use when a user asks to find, compare, or install a skill, asks whether a skill exists for a task, or wants to extend agent capabilities with a reusable workflow.

yenhao-huang Bundle

Download complete Hugging Face model repositories or selected files into ~/Desktop/models through the existing HF CLI environment, using tmux for observable long-running transfers and updating the local model inventory. Use when the user asks to download, resume, or verify model weights locally.

yenhao-huang Bundle

Query the connected Dart task-management MCP and produce the daily Traditional Chinese status summary for today, rag, model-tester, pretrievalagent, and esg contest. Use for the migrated daily Dart cron, manual task briefings, or reruns of a missing summary.

yenhao-huang Bundle

Rank this month's newly published Hugging Face text-generation, text-to-image, and text-to-audio models by likes and save a Traditional Chinese Top 5 survey for each category. Use for the migrated weekly model cron, a manual weekly model survey, or a missing report rerun.

uwayxt Bundle

Design intuitive UX patterns for AI interfaces including streaming, confidence, overrides, and prompts. Use when AI features, prompt UX, streaming UI, LLM interface, or AI assistant.

uwayxt Bundle

Translate a Business Model Canvas into product modules. Use when bridging business strategy to product architecture.

uwayxt Bundle

Evaluate monetization and unit economics. Use when how will this make money, business model, monetization, unit economics, or pricing.

nemori-ai Bundle

Use when you (a coding/execution agent) are designing, building, or testing actual code and want to apply established engineering theories well —— 当你(执行 agent)在动手设计 / 开发 / 测试代码、想遵循 DDD / SDD / TDD / OOP 的最佳实践与品味时:怎么给领域建模(bounded context / aggregate / 不变式)、怎么写好类(SOLID / 四柱 / 组合优于继承)、怎么 spec-first(契约即 SSOT)、怎么 test-first(红绿铁律 / constraint parity)。核心是工程心智模型 + 哲学 + 艺术品味 + 方法论,叠加 no-silent-failure / spec 不漂移 / 无 idle wrapper 等红线。五条共享根:不变式即锚 / 契约即 SSOT / 组合优于包装 / 分层思维 / 证据优于声称。Triggers: 要设计一个模块 / 给领域建模 / 划 bounded context / 写类纠结继承还是组合 / 要不要先写 spec / 先写测试还是先写实现 / 这个抽象值不值 / 怎么躲开贫血模型或 god class。Do NOT use when 你在决定该编排什么 / 怎么派发(master-orchestrator-guide)、怎么把目标切成任务 DAG(slicing-goals-into-dags)、把一个任务迭代优化到验收的循环形状本身(dev-as-ml-loop)、怎么用 ccm 写 board(using-ccm)、workflow 脚本怎么写(authoring-workflows)。

grasscaograss Bundle

Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.

grasscaograss Bundle

完整的 agent 开发方法论技能集。涵盖从头脑风暴到分支完成的完整工作流：设计→计划→实现→审查→交付。当需要结构化开发流程、TDD、系统调试、代码审查、git worktree 隔离开发时使用。源自 Jesse Vincent 的 superpowers 项目。

grasscaograss Skill

Import external documents into the project archive. Two modes: → records (format conversion with TL;DR) for specs and reports, or → knowledge (multi-agent parallel organization with dedup/merge) for technical references and wiki dumps. Use when user wants to import external docs, zread output, spec documents, or reference materials into the archive.

cookiemonsterdev Bundle

Use when users ask how to install, configure, model, query, index, secure, scale, repair, compact, back up, monitor, or troubleshoot Apache Cassandra clusters and CQL applications, including keyspaces, tables, partition keys, clustering columns, SAI, secondary indexes, replication, compaction, nodetool, cqlsh, vector search, CDC, or the native protocol, especially when they need official Apache Cassandra documentation, exact links, or docs-grounded guidance.

cookiemonsterdev Bundle

Explains and applies the database normal forms (1NF, 2NF, 3NF, BCNF, 4NF, 5NF) and strategic denormalization, with worked examples in PostgreSQL, MySQL, SQLite, MongoDB, Cassandra, ScyllaDB, and DynamoDB. Use when users ask about normal forms, functional/multivalued/join dependencies, anomalies (insertion/update/deletion), schema design and refactoring, when they need help deciding how far to normalize a schema, or when they want a normalization concept translated into a specific database's modeling idioms (relational tables, MongoDB embedding/referencing, Cassandra/ScyllaDB query-driven tables, DynamoDB single-table design).

griddynamics Bundle

To build an AI harness: run, observe, validate, automate repeated work faster — CLI/MCP actions, devcontainers, skills, subagents, hooks, pipelines, automations.

griddynamics Bundle

To delegate, prompt and manage subagents. MUST activate to spawn subagent with a quality prompt, unless trivial.

griddynamics Bundle

To connect Rosetta with Grid Dynamics SpecFlow MCP; only when SpecFlow is mentioned and the MCP is installed.