Altertable Skills

CI Score Skills Agent Skills License: MIT Skills Python 3.12+

A collection of Agent Skills for Altertable AI agents. Skills are portable, version-controlled packages of instructions, scripts, and references that give agents new capabilities.

What are Agent Skills?

Agent Skills are an open standard for giving AI agents specialized capabilities. Each skill is a self-contained folder with a SKILL.md file containing instructions that agents load on demand. Build once, use across any compatible platform.

Quick Start with /altertable:ask

Don't know which skill to use? Just type:

/altertable:ask <your question>

The /altertable:ask command routes your query to the best skill automatically. Examples:

  • /altertable:ask show me my web traffic -> analyze-web-traffic
  • /altertable:ask what tables do I have? -> explore-data
  • /altertable:ask analyze my signup funnel -> analyze-funnels
  • /altertable:ask what is Altertable? -> understand-platform

Available Skills

Skill Description
analyze-funnels Creates and analyzes conversion funnels to understand user journeys
analyze-insights Interprets chart data to identify patterns, anomalies, and trends
analyze-web-traffic Analyzes web analytics data to identify traffic patterns
build-segments Creates user segments and cohorts using filters and dimensions
configure-tasks Configures autonomous background tasks for AI analysis
create-insights Creates discoveries with insights through the approval workflow
create-discoveries Creates discoveries from meaningful findings that should notify users and enter review
decide-actions Decision matrices for choosing insight types and discovery actions
evaluate-skills Evaluates and creates agent skills following best practices
explore-data Explores data connections and schemas
forecast-timeseries Analyzes time series data for trends, anomalies, and forecasts
instrument-product-analytics Adds product event and identity instrumentation to applications
manage-discoveries Manages the discovery approval workflow and user feedback
query-lakehouse Writes and executes SQL queries against the DuckDB Lakehouse
query-product-events Queries product events, identities, traits, and instrumentation delivery
understand-platform Explains platform concepts and architecture
use-memory Stores and retrieves agent memories for learning and context

Getting Started

git clone https://github.com/altertable-ai/skills.git
cd skills
uv sync
uv run pre-commit install

Validate a skill

uv run skills validate ./skills/skill-name

Score a skill

Every skill is scored by an LLM judge against the Agent Skills spec (threshold: 70/100):

uv run python scripts/score-skills.py ./skills/skill-name --verbose

Run tests

uv run pytest scripts/tests/ -v

Creating a New Skill

cp -r skills/SKILL_TEMPLATE skills/my-new-skill

Each skill follows the Agent Skills Specification:

skills/
  skill-name/
    SKILL.md          # Required: metadata + instructions
    references/       # Optional: detailed documentation
    scripts/          # Optional: executable code
    assets/           # Optional: templates, resources

See CONTRIBUTING.md for full guidelines.

Compatibility

Skills follow the open Agent Skills standard and work across 30+ compatible platforms:

Claude Code | Cursor | VS Code | Gemini CLI | OpenAI Codex | GitHub Copilot | Goose | Roo Code | OpenHands | and more

Resources

License

This project is licensed under the MIT License. See LICENSE for details.