olakai

Olakai Skills

Official skills for integrating AI agents with Olakai - the enterprise AI observability and governance platform.

These skills follow the Agent Skills Standard and work with Claude Code, Cursor, VS Code Copilot, and other compatible AI coding assistants.

Installation

Using add-skill (Recommended)

# List available skills
npx add-skill olakai-ai/olakai-skills --list

# Install a specific skill
npx add-skill olakai-ai/olakai-skills/olakai-get-started
npx add-skill olakai-ai/olakai-skills/olakai-new-project
npx add-skill olakai-ai/olakai-skills/olakai-integrate
npx add-skill olakai-ai/olakai-skills/olakai-troubleshoot
npx add-skill olakai-ai/olakai-skills/olakai-reports

Manual Installation

# Clone to user-level skills directory
git clone https://github.com/olakai-ai/olakai-skills ~/.claude/skills/olakai-skills

# Or clone to project-level
git clone https://github.com/olakai-ai/olakai-skills .claude/skills/olakai-skills

Available Skills

Skill	Description
olakai-get-started	Install the CLI, authenticate, and send your first monitored event
olakai-new-project	Build a new AI agent project from scratch with full observability
olakai-integrate	Add Olakai to existing AI code with minimal changes
olakai-troubleshoot	Diagnose and fix issues with events, KPIs, or SDK integration
olakai-reports	Generate usage summaries, KPI trends, ROI reports from the terminal

Bundled Agent

The olakai-expert agent combines all four skills into a single specialist:

Agent	Description
olakai-expert	Full Olakai integration specialist - creates agents, adds monitoring, troubleshoots issues, generates reports

Prerequisites

Before using these skills, ensure you have:

1. Olakai CLI installed: npm install -g olakai-cli
2. CLI authenticated: olakai login
3. API key for SDK integration (generated per-agent via CLI)

The Golden Rule: Test - Fetch - Validate

Always validate your Olakai integration by generating a test event and inspecting it:

# 1. Run your agent/trigger LLM call
# 2. Fetch the event
olakai activity list --agent-id AGENT_ID --limit 1 --json
olakai activity get EVENT_ID --json | jq '{customData, kpiData}'

# 3. Validate:
#    - customData has all expected fields
#    - kpiData shows NUMBERS (not strings like "MyVariable")
#    - kpiData shows VALUES (not null)

Quick Reference

# CLI Authentication
olakai login
olakai whoami

# View Activity
olakai activity list --limit 10
olakai activity get EVENT_ID --json
olakai activity sessions --agent-id ID  # Session decoration status

# Manage Agents (includes API key generation)
olakai agents list
olakai agents create --name "Agent Name" --with-api-key --json
olakai agents get AGENT_ID --json | jq '.apiKey'

# Manage KPIs
olakai kpis list --agent-id ID
olakai kpis create --formula "Variable" --agent-id ID

# Manage Custom Data
olakai custom-data list
olakai custom-data create --name "Field" --type NUMBER

SDK Quick Start

TypeScript:

import { OlakaiSDK } from "@olakai/sdk";
const olakai = new OlakaiSDK({ apiKey: process.env.OLAKAI_API_KEY! });
await olakai.init();
const openai = olakai.wrap(new OpenAI({ apiKey }), { provider: "openai" });

Python:

from olakaisdk import olakai_config, instrument_openai
olakai_config(os.getenv("OLAKAI_API_KEY"))
instrument_openai()

Optional: Improved Skill Discovery

For automatic skill activation (recommended for heavy Olakai users):

# Install activation hooks for improved auto-invocation rate
cp hooks/skill-activator.sh ~/.claude/hooks/
chmod +x ~/.claude/hooks/skill-activator.sh

See hooks/README.md for full setup instructions.

Documentation

• Olakai Documentation
• TypeScript SDK
• Python SDK
• CLI Reference

Agent Compatibility

See AGENTS.md for full compatibility information with various AI coding agents.

License

MIT License - see LICENSE for details.

Contributing

Contributions welcome! Please read our contributing guidelines before submitting PRs.

---

Built with love by Olakai