AI & ML

oliver-kriska Bundle

Recall prior work from past sessions — how a bug was fixed, what was decided, where a pattern lives. Use when asked 'have we done this before' or 'how did I fix X' in Elixir/Phoenix work. ccrider MCP when available, else git + solution docs.

oliver-kriska Bundle

Fix Elixir/Phoenix code until Codex CLI review comes back clean — bounded review, fix, verify loop before opening a PR. Use when codex is installed and you want an external cross-model critic on your changes before pushing.

tinyhumansai Skill

Build a model whose assumptions are enumerable and whose conclusion can be tested against them.

jamie-bitflight Skill

Use when checking skill quality, validating frontmatter before commit, or diagnosing validator warnings. Runs the plugin validator on a skill, agent, or plugin directory — reports token complexity, broken links, frontmatter issues, and structural problems. Pass the path as an argument.

jamie-bitflight Skill

Read The Fucking Prompt — finds the strongest user reaction to an AI instruction-following failure in a chosen session, reconstructs what the assistant did wrong, and renders a shareable terminal-style PNG. Use when asked to find rage moments, generate a rage receipt, or capture a frustration incident from a session.

jamie-bitflight Skill

Use when the primary outcome is refining an existing AI-facing artifact without dropping behavior, including SKILL.md, AGENTS.md, CLAUDE.md, rules, prompts, and agent definitions that need sharper invocation, structure, or completion criteria.

sprintberlin Bundle

Zoho WorkDrive via MCP with action catalog, least-privilege profiles, team, folder, file, search, and share helper scripts, and verified document workflows.

caezium Skill

Diagnose and fix the user's Mac with Burrow's local MCP tools (burrow_doctor, burrow_snapshot, burrow_top_processes, burrow_process_usage, burrow_ports, burrow_analyze, burrow_disk_forecast, burrow_dupes, burrow_anomalies, burrow_agent_audit, burrow_clean, …). Use whenever the Mac is slow, hot, loud, low on disk, draining battery, or misbehaving; when the user asks what's using CPU/memory, what's listening on a port, what's eating disk space, where the duplicate or leftover files are, whether anything is behaving unusually, or what an agent already changed; AND proactively — if you notice a system problem mid-task (low disk, a runaway process, a port conflict), reach for these tools to diagnose and offer a fix without being asked. Requires Burrow's MCP server connected (burrow_* tools available).

jtapes Bundle

File-based message bus between Claude Code agents (projects, subagents and the user through a local web UI) with messages, tasks, background wake-ups and cron schedule. Use on "ask X", "have X do", "tell X", "what is in the inbox", "open the bus UI", "create agent X", "connect this project to the bus", "run on a schedule". Файловая шина между агентами Claude Code — проектами, субагентами и пользователем (веб-UI) — сообщения, задачи, подъём получателя. Используй на «спроси у X», «пусть X сделает», «передай / сообщи X», «шо во входящих», «переписка с X», «открой шину», «создай агента X», «подключи проект к шине», «запускай по расписанию / по крону».

jamie-bitflight Skill

Use when orchestrating a Python development task via specialized agents. Activates on "build a Python CLI", "add a feature", "write tests", "refactor Python code", "debug Python", "code review", or any multi-agent Python workflow. Invoke as /orchestrate with a task description or alone to use conversation context.

thobai Bundle

Ask Thomas something in Google Chat without blocking. His reply arrives as a new prompt in this pane.

2710074390-cyber Bundle

医学题库生产管线（五阶段 Agent 工作流：出题→质检→修复→成册），带确定性门禁校验、Bloom 认知分层配额、金标准比对与回归教训库。Invoke when 用户要从教材/笔记批量生成医学题库或复习资料、提到 MedAgentWork、题库生产、出题质检、押题卷、复习手册生成，或需要对已有题库做质量门禁校验。

jamie-bitflight Skill

Use as the routing layer for Python development tasks — matches task descriptions against trigger lists and activates specialist skills before starting work. Covers Typer, Rich, Textual, FastMCP/MCP, ty type checker, uv, Hatchling, TOML editing, pre-commit/prek, async Python, PyPI packaging, complex linting, and technical debt modernization.

emadmokhtar Bundle

Use when writing, running, or auditing skill-lens eval suites for an Agent Skill — deciding which cases a skill needs, choosing between assertions, judge rubrics and trajectory checks, and reading a failing case correctly

ai-analyst-lab Skill

Causal inference toolkit for when experiments are not possible: estimate treatment effects from observational data with assumption checks and mandatory caveats. Invoke as /causal. Trigger on "causal", "caused", "impact of", "effect of", "attribution", "counterfactual", "difference-in-differences", "DiD", "propensity matching", "pre-post". If randomization IS possible, route to /experiment design instead.

ai-analyst-lab Skill

Browse and search past analyses from the knowledge system's analysis archive. Use this skill whenever the user invokes `/history` or variants like `/history search=X`, `/history {id}`, `/history --all`, `/history dataset=X`, or asks questions about their analytical work history such as "what have I analyzed before?", "what analyses have I run?", "show me past work", "show my history", "what have we looked at?", "what questions have I answered?", "can I see my analysis history?", "list my analyses", "what did I work on last week?", "have I analyzed X before?", "find analyses about Y", or mentions reviewing prior analyses, checking if similar work was already done, needing context on previous findings, or wants to build on past work. This skill also applies automatically at session start when you need context on the user's analytical history to inform current work. When displaying history, ALWAYS filter to the active dataset unless the user explicitly requests --all or dataset={id}.

ai-analyst-lab Skill

Browse, search, and display metric definitions from the active dataset's metric dictionary. This skill provides quick access to how metrics are defined, computed, and validated. Use this skill whenever the user wants to see metric definitions, understand how a metric is calculated, check what metrics are available, or verify a metric's specification before using it in analysis. Trigger on phrases like "/metrics", "show me the metrics", "what metrics do we track?", "how is [metric name] calculated?", "what's the definition of [metric]?", "list all metrics", "show me revenue metrics", "what metrics are in this dataset?", "define conversion rate", "how do we measure retention?", "what's in the metric dictionary?", "search for engagement metrics", "show me all KPIs", or any question about metric definitions, specifications, or formulas. Also use this skill during analysis when you need to confirm a metric's exact definition before computing it, or when the user references a metric name and you need to verify its

ai-analyst-lab Skill

Produce a comprehensive cross-table data health map for the active dataset — the full payoff answer to open-ended "tell me about this data" questions. Run all tables, not just one: table inventory with row counts, PK uniqueness, date range per table, date alignment across tables, column completeness, foreign-key join-rate matrix, relationship diagram, and a thread-to-pull opening hypothesis. Use this skill whenever the user asks a **dataset-wide open question** — "tell me about this data", "tell me about the data", "tell me about this dataset", "what's in here", "what's in this data", "give me an overview", "give me the map", "map out the data", "what do I have", "what does this data look like", "show me what we've got", or any broad first-contact question about the active dataset as a whole (not scoped to one table). This skill is the curriculum payoff moment — students who just connected data get cross-table health, relationship mapping, and date alignment on the first broad question, not a schema dump and

ai-analyst-lab Skill

List all connected datasets with their status, table counts, and last analysis date. Use this skill whenever the user invokes `/datasets`, or asks questions like "what datasets do I have?", "show me my data sources", "list datasets", "which datasets are connected?", "what data is available?", "show all datasets", "what datasets can I analyze?", "view my datasets", "what data sources are set up?", or any request to see what datasets exist in the system. Also trigger when users mention "switch dataset", "change dataset", "use a different dataset", or "what's my active dataset?" since seeing the list helps them choose. This is a foundational command that should be offered proactively whenever users seem unsure about what data they're working with or when they need to understand what datasets are available before starting analysis.

ai-analyst-lab Skill

Automatically detect Sample Ratio Mismatch (SRM) in experiment or A/B test data before any analysis proceeds. SRM is a critical randomization integrity check — if the treatment/control split deviates significantly from the expected ratio, the experiment is compromised and results cannot be trusted. This skill acts as a safety gate that blocks analysis when randomization is broken. Use this skill whenever you detect experiment or A/B test data — look for columns like "variant", "group", "treatment", "control", "arm", "experiment_group", "test_group", "bucket", "condition", or any column with binary/small-cardinality values that suggest treatment assignment. Auto-fire on experiment data detection without waiting to be asked. Apply this skill when loading any experiment dataset, before calculating treatment effects, when starting any experiment analysis workflow, when users mention "A/B test", "experiment", "treatment vs control", "randomization", "test group", or when you see data that looks like it came from a

ai-analyst-lab Skill

Show the active dataset's schema — tables, columns, row counts, and relationships. Optionally drill into a specific table. Use this skill whenever the user invokes `/data` or `/data {table}`, or asks questions like "what tables do I have?", "show me the schema", "what's in this dataset?", "what columns are in the users table?", "show me table structure", "list tables", "describe the data", "what's in my database?", or any request to inspect, browse, or understand the structure of the active dataset. Also trigger when users mention "schema", "columns", "tables", "data dictionary", "data structure", or when they need to understand what data they're working with before starting an analysis. This is a foundational command that should be offered proactively when users seem unsure about available data or table structure. DISAMBIGUATION: this is the `/data` SCHEMA inspector (structure — tables, columns, types, row counts). For a dataset-wide health + relationships overview ("tell me about this data"), use `data-map`

ai-analyst-lab Skill

Connect Claude Code to Notion's official hosted MCP, complete OAuth, verify the intended workspace with a read, and stop before any write. Use when the user asks to connect Notion, set up Notion, export to Notion, or diagnose an unavailable Notion connection.

deadmade Bundle

Use for "interrogate", "adversarial review", "multi-model review", "challenge this", "stress test this code", "find blind spots", or "tear this apart". Multiple LLM reviewers challenge changes from independent angles.

stanfish06 Skill

Building ML demos and web UIs in Python with Gradio 6 — gr.Interface for wrapping a function, gr.Blocks for custom layouts with event listeners, gr.ChatInterface for LLM chat, streaming generator outputs, gr.State, image/audio components, queueing and concurrency, launch()/share links, and mounting into FastAPI with gr.mount_gradio_app. Use when demoing a model (image, audio, text, LLM), building a quick UI around a Python function, sharing a prototype via a public link, or hosting on Hugging Face Spaces.

stanfish06 Bundle

Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing repositories, models, datasets, and Spaces on the Hugging Face Hub. Replaces now deprecated `huggingface-cli` command.

stanfish06 Bundle

Full-featured computational pathology toolkit. Use for advanced WSI analysis including multiplexed immunofluorescence (CODEX, Vectra), nucleus segmentation, tissue graph construction, and ML model training on pathology data. Supports 160+ slide formats. For simple tile extraction from H&E slides, histolab may be simpler.

ai-analyst-lab Skill

Publish an approved analysis as a Notion text page through the official hosted MCP. Use when the user asks to export, publish, create, share, put, or send analysis results to Notion.

ai-analyst-lab Skill

Assess whether the data needed for an analysis actually exists, identify what's missing, and produce prioritized instrumentation requests for engineering when gaps are found. Use this skill whenever you're about to start an analysis, when a user asks about data availability, when you need to check if certain events or properties are tracked, when the Data Explorer agent finds incomplete data, when you're designing an analysis and need to verify data exists, when initial queries return nulls or missing values, when a user mentions "do we track...", "is there data for...", "can we measure...", when you're writing an instrumentation request, when you need to assess data completeness, when planning metrics or experiments that require specific tracking, when a user wants to know what data gaps exist, or any time you need to map analytical requirements to available data sources. This skill should be applied proactively before committing to an analytical approach to avoid wasted work on analyses that can't be comple

ai-analyst-lab Skill

Verify Google Workspace MCP authentication at the start of any session that needs Google APIs (Docs, Slides, Drive). This skill prevents auth failures mid-workflow by testing credentials upfront. Use this skill automatically at session start when the task involves Google Docs, Google Slides, Drive uploads, or any MCP Google Workspace tool. Also trigger when users mention "Google Doc", "Google Slides", "upload to Drive", "export to Google", "share on Drive", or any Google-related output format. Apply before running any Google Workspace agents (google-slides-creator, google-slides-reviewer) or calling any mcp__google-* tool. This skill detects the actual MCP configuration, checks stored credentials in all known locations, tests tokens with a lightweight API call using create operations instead of reads, handles re-authentication if needed, and reports auth status clearly so downstream work can proceed safely or fail fast with actionable guidance.

ai-analyst-lab Skill

Deep-profile the active dataset: distributions, temporal patterns, correlations, completeness gaps, anomalies. Use after connecting a new dataset. Trigger on "profile this data", "deep-profile the dataset", "run a data profile", "check distributions", "find anomalies in the data", "how complete is this data". For a first-contact overview use data-map; for one column use distribution-profiler; for schema use data-inspect.

deadmade Skill

Apply when writing stateful logic, or when code branches a lot or repeats a shape assumption across files. Encode the domain in a structure instead of scattered conditionals.

stanfish06 Bundle

RNA velocity analysis with scVelo. Estimate cell state transitions from unspliced/spliced mRNA dynamics, infer trajectory directions, compute latent time, and identify driver genes in single-cell RNA-seq data. Complements Scanpy/scVI-tools for trajectory inference.

ai-analyst-lab Skill

Compare a narrow model grader with frozen human labels and inspect disagreement, bias probes, and repeated scoring stability. Use when building or changing a model-based evaluator.

ai-analyst-lab Skill

Compare metrics, findings, and patterns across two or more connected datasets. Helps identify cross-dataset patterns (e.g., "conversion funnel behavior is similar across both product lines") and dataset-specific anomalies. Use this skill whenever the user wants to compare datasets, mentions phrases like "compare datasets", "is this pattern the same across", "how does X differ between datasets", "cross-dataset comparison", "analyze across product lines", "compare metrics between", or asks whether a finding is unique to one dataset or appears universally. Also trigger when the user has analyzed multiple datasets sequentially and might benefit from seeing commonalities and divergences side-by-side. If you've just finished analyzing dataset A and the user switches to dataset B to run similar queries, proactively offer to compare the two. This skill is valuable for identifying where business patterns are universal vs. dataset-specific, spotting metric definition inconsistencies, and finding opportunities to apply

ai-analyst-lab Skill

Create properly formatted Google Docs via the MCP API. This skill prevents common issues like text/image overlap, broken heading hierarchy, excessive whitespace, and inconsistent formatting. Use this skill automatically whenever you're building a Google Doc, calling any Google Docs MCP tool on the google-workspace server (create_doc, insert_doc_elements, insert_doc_image, batch_update_doc) or the google-docs server (upload_file_to_drive, write_formatted_content), designing a document structure, or when the google-doc-creator or google-doc-reviewer agent is running. This skill is essential for ANY workflow involving Google Docs creation, document formatting, analysis writeups in Google Docs, report generation to Docs, chart embedding in documents, or exporting analysis results to shareable Docs. Make sure to use this skill whenever the user wants to create a Doc, export to Google Docs, share analysis as a Doc, build a formatted document, or mentions Google Docs in any capacity.

jmix-framework Skill

Before typing a class, enum constant, inner-class event, method, or icon name you have not personally verified in this project, confirm it exists. Costs seconds and prevents the most expensive class of failures (hallucinated icon constants, fake event inner classes, wrong package paths). Primary check when connected = the Context7 docs MCP (/jmix-framework/jmix-context7); falls back to grepping the project for a working example or an IDE symbol search.