# AI Agent Skills

Extend your AI coding agent with Kopai observability skills

Skills are portable instruction files that extend AI coding agents with domain-specific knowledge. Instead of reading docs and explaining best practices to your agent manually, skills give your agent direct awareness of Kopai's CLI, workflows, and conventions.

Kopai ships three skills that cover the full observability workflow — from instrumenting your app, to debugging production issues, to building dashboards.

## What are skills?

Skills follow the [Agent Skills open standard](https://agentskills.io/) — `SKILL.md` files with YAML frontmatter that any compatible agent can read. Your agent loads the skill's instructions and automatically invokes them when your conversation matches the skill's description, or when you trigger them manually via slash command.

The standard is supported by [20+ AI coding agents](https://skills.sh/) including [Claude Code](https://docs.anthropic.com/en/docs/claude-code/skills), [Cursor](https://docs.cursor.com/context/skills), and [Codex CLI](https://github.com/openai/codex). See the [Agent Skills specification](https://agentskills.io/specification) for details.

## Install Kopai skills

Pick one of the three options below. Each installs the same skills.

**Option 1 - skills CLI** (all three skills at once, pulls latest from GitHub):

**npm:**

```bash
npx skills add kopai-app/kopai-mono
```

**pnpm:**

```bash
pnpm dlx skills add kopai-app/kopai-mono
```

**yarn:**

```bash
yarn dlx skills add kopai-app/kopai-mono
```

**Option 2 - Tessl registry** (install individually, version-pinned, quality-reviewed):

**npm:**

```bash
npx tessl i kopai/otel-instrumentation
npx tessl i kopai/root-cause-analysis
npx tessl i kopai/create-dashboard
```

**pnpm:**

```bash
pnpm dlx tessl i kopai/otel-instrumentation
pnpm dlx tessl i kopai/root-cause-analysis
pnpm dlx tessl i kopai/create-dashboard
```

**yarn:**

```bash
yarn dlx tessl i kopai/otel-instrumentation
yarn dlx tessl i kopai/root-cause-analysis
yarn dlx tessl i kopai/create-dashboard
```

Browse all Kopai skills on the [Tessl registry](https://tessl.io/registry/kopai).

**Option 3 - manually**, if you prefer not to use a package manager: clone the [kopai-mono repo](https://github.com/kopai-app/kopai-mono) and copy the `skills/` directories into your agent's skills folder. See the docs links above for your agent's exact path.

| Agent | Skills directory | Docs |
|-------|-----------------|------|
| [Claude Code](https://code.claude.com/docs/en/skills#where-skills-live) | `.claude/skills/` | [Where skills live](https://code.claude.com/docs/en/skills#where-skills-live) |
| [GitHub Copilot](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/create-skills) | `.github/skills/` | [Creating skills](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/create-skills) |
| [Cursor](https://cursor.com/docs/context/skills) | `.cursor/skills/` | [Skill directories](https://cursor.com/docs/context/skills) |
| [Codex CLI](https://developers.openai.com/codex/skills#where-to-save-skills) | `.codex/skills/` | [Where to save skills](https://developers.openai.com/codex/skills#where-to-save-skills) |
| [Gemini CLI](https://geminicli.com/docs/cli/skills/#skill-discovery-tiers) | `.gemini/skills/` | [Skill discovery tiers](https://geminicli.com/docs/cli/skills/#skill-discovery-tiers) |
| [OpenCode](https://opencode.ai/docs/skills/#place-files) | `.opencode/skills/` | [Place files](https://opencode.ai/docs/skills/#place-files) |

:::note
Review the skill files before installing. Skills provide instructions that shape how your agent works — read through them to ensure you're comfortable with the content.
:::

## Available skills

### `otel-instrumentation`

Instruments your application with OpenTelemetry SDK and validates that telemetry reaches Kopai.

- **Supports:** Node.js, Next.js, Python, Go, Java, .NET
- **Triggers automatically when you say things like:** "add observability to my app", "set up tracing", "my traces aren't showing up"
- **Slash command:** `/otel-instrumentation`
- **Prerequisite:** Kopai backend running (`npx @kopai/app start`)
- **What to expect:** your agent installs the right OTEL SDK packages, creates the instrumentation setup file, configures environment variables, and validates data is flowing using `npx @kopai/cli traces search`

[View on GitHub →](https://github.com/kopai-app/kopai-mono/tree/main/skills/otel-instrumentation)

### `root-cause-analysis`

Analyzes telemetry data to debug errors, latency issues, and distributed failures.

- **Triggers automatically when you say things like:** "why is my API slow", "getting 500 errors", "service is down", "requests are timing out"
- **Slash command:** `/root-cause-analysis`
- **Prerequisite:** services sending OpenTelemetry data to Kopai
- **What to expect:** your agent searches error traces → gets full trace context → correlates logs → checks metrics → presents findings with a summary, evidence, impact assessment, and suggested fix

[View on GitHub →](https://github.com/kopai-app/kopai-mono/tree/main/skills/root-cause-analysis)

### `create-dashboard`

Builds observability dashboards from OTEL metrics, logs, and traces.

- **Triggers automatically when you say things like:** "create a dashboard", "show me a metrics view", "build a monitoring panel"
- **Slash command:** `/create-dashboard`
- **Prerequisite:** metrics data ingested into Kopai
- **What to expect:** your agent discovers available metrics via `npx @kopai/cli metrics discover`, designs a component layout, and creates the dashboard via the CLI — you get a link to view it in the Kopai web UI

[View on GitHub →](https://github.com/kopai-app/kopai-mono/tree/main/skills/create-dashboard)