datadog.nvim

A Neovim plugin for browsing Datadog Error Tracking issues for the current git repository — list issues on the left, inspect each one (stack trace, first/last seen, commit hashes) in a side detail pane, and jump straight to the offending line in your code.

Features

• Bottom split with issue list (left) and detail pane (right) — cursor on the list updates the detail view as you move
• Per-issue detail shows: error type & message, service/env/host, first seen and last seen times with the commit hashes (first_seen_version / last_seen_version) where the issue was first and most recently observed, total occurrences, an inline occurrence sparkline (▁▂▅▇█▆▃) covering the configured time range, and the full stack trace
• <CR> on any stack-frame line jumps to that file:line (works for Java, Go, JS, Ruby, Rust, Kotlin, Python frames out of the box)
• Issues are scoped to the current git repository (@git.repository.id filter), with optional service/env filters

Installation

Using packer.nvim

use {
  'teekey-dev/datadog.nvim',
  requires = {
    'nvim-lua/plenary.nvim', -- For HTTP requests
    'MunifTanjim/nui.nvim',    -- For popup UI
    'nvim-web-devicons',     -- Optional: for file icons
  },
  config = function()
    require('datadog').setup({
      api_key = 'your-datadog-api-key',
      app_key = 'your-datadog-app-key',
      -- Optional configuration
      site = 'us1', -- Datadog site (us1, us3, us4, us5, eu1, gov)
      refresh_interval = 30000, -- 30 seconds in milliseconds
      query = {
        time_range = '1h', -- Time range for queries (e.g., '30m', '1h', '24h')
        filter = {
          query = '@status:error', -- Datadog log query
        },
        limit = 100, -- Maximum number of errors to fetch
      },
    })
  end
}

Using lazy.nvim

{
  'teekey-dev/datadog.nvim',
  dependencies = {
    'nvim-lua/plenary.nvim',
    'MunifTanjim/nui.nvim',    -- For popup UI
    'nvim-web-devicons', -- Optional
  },
  opts = {
    api_key = 'your-datadog-api-key',
    app_key = 'your-datadog-app-key',
    -- Add other configuration options as needed
  }
}

Usage

Once installed and configured, use the following commands:

• :DatadogErrors - Open the Datadog errors buffer
• :DatadogRefresh - Manually refresh the error data
• :DatadogConfig - Show the current plugin configuration

In the errors buffer:

• On the list pane (left):
  • Move with j / k (or any motion) — the detail pane updates as the cursor moves
  • <Enter> - Focus the detail pane so you can browse the stack trace
  • r - Refresh the error list
  • q or <ESC> - Close the errors buffer
• On the detail pane (right):
  • j / k - Scroll the detail / stack trace
  • <Enter> - Jump to file:line under the cursor (works on stack-frame lines); falls back to the issue's primary file if the line has no frame
  • q or <ESC> - Close the errors buffer

If the terminal is shorter than 25 lines, the plugin falls back to a list-only view automatically.

Configuration

The plugin can be configured with the following options:

Option	Description	Default
api_key	Your Datadog API key	"" (required)
app_key	Your Datadog application key	"" (required)
site	Datadog site (us1, us3, us4, us5, eu1, gov)	"us1"
refresh_interval	Automatic refresh interval in milliseconds	30000 (30 seconds)
query.time_range	Time range for log queries (e.g., '1h', '24h')	"1h"
query.filter.query	Datadog log query string	"@status:error"
query.limit	Maximum number of logs to fetch	100

How It Works

1. On open, the plugin runs git remote get-url origin to find the current repository URL and converts it to a Datadog @git.repository.id:github.com/owner/repo filter.
2. It calls POST /api/v2/error-tracking/issues/search with that filter (plus any optional service / env filters and the configured time range) to get the issues for this repo, including first_seen / last_seen timestamps and first_seen_version / last_seen_version commit hashes.
3. In parallel:
   • POST /api/v2/spans/events/search (one batched request OR'd over all issue IDs) fetches a representative span per issue carrying the file/line, error message, and full stack trace.
   • POST /api/v2/spans/analytics/aggregate returns a count timeseries grouped by @issue.id, used to render the per-issue occurrence sparkline. This endpoint is rate-limited to 300 requests/hour by Datadog; if it fails (rate limit, etc.) the trend line shows (unavailable) and the rest of the detail still renders.
4. The responses are merged and rendered: the left pane shows the issue table; the right pane shows the detail of the selected issue, updated live as the cursor moves.
5. <Enter> on a stack-frame line in the detail pane parses file:line from that line and opens the file at the matching line.

Requirements

• Neovim >= 0.5.0
• plenary.nvim (for HTTP requests)
• Datadog API and Application keys

Development

This plugin follows Neovim Lua plugin best practices:

• Modular architecture with separate concerns (API, buffer management, utilities)
• Functional programming principles where appropriate
• Proper error handling and validation
• Clear separation of plugin state and configuration

License

MIT