skills.rest
latitude-dev
latitude-devOFFICIAL
Explore all skills by this author →
Website𝕏@trylatitudeGitHub

fix-datadog-issues

Triage and fix production errors from Datadog Error Tracking, then open a pull request.

4.6k|387|Updated Jun 28, 2024
Search Tags:#root cause#triage#datadog#Debugging#error tracking#production errors#Pull request
One-click install
npx skills add https://github.com/latitude-dev/latitude-llm --skill fix-datadog-issues
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fix-datadog-issues
Source: https://github.com/latitude-dev/latitude-llm/tree/main/.agents/skills/fix-datadog-issues
Command: npx skills add https://github.com/latitude-dev/latitude-llm --skill fix-datadog-issues

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Production errors pile up in Datadog Error Tracking with no systematic way to triage them, separate real code bugs from noise, and turn them into reviewed fixes. This Skill walks an agent from querying Datadog issues through root-cause analysis to a tested fix and a pull request.

Core Features & Use Cases

  • Issue discovery and triage: Query Datadog Error Tracking via the Datadog MCP server (or span aggregation fallback), classify candidates as code bugs versus infra noise, version skew, or expected errors, and rank by impact, trend, and fix confidence.
  • Root-cause analysis: Map error spans to code using service names, resource names, stack traces, and service.version commit SHAs, then identify the introducing commit with git blame and deploy correlation.
  • Test-first fixing and PR creation: Reproduce the bug with failing tests, apply the fix at the correct architectural layer, verify with typecheck and lint, comment on the Datadog issue, and open a PR against the development branch.
  • Use Case: Ask the agent to investigate the most frequent production errors this week; it will filter out transient timeouts and expected errors, pick a genuine bug, confirm a fix plan with you, then deliver a tested PR linked to the Datadog issue.

Quick Start

Ask the agent to find the most frequent production errors in Datadog Error Tracking from the last week and fix the top genuine code bug.

Frequently Asked Questions about fix-datadog-issues

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find and fix production errors from Datadog?▼

Query Datadog Error Tracking issues or aggregate error spans filtered by env:production, classify candidates to separate genuine code bugs from infra noise, then root-cause the top issue in code. Reproduce it with a failing test, apply the fix, and open a PR against the development branch.

How do I query Datadog Error Tracking issues with the MCP server?▼

Use the Datadog MCP server (plugin:datadog:mcp) with the error-tracking toolset enabled, discovering tool names via ToolSearch. If the toolset is off, fall back to aggregate_spans and search_datadog_spans, which stamp issue IDs onto error spans via custom.issue.id.

Does this work with Datadog Incident Management incidents?▼

No. The Skill targets Error Tracking issues, which are fingerprinted groups of error occurrences. Datadog Incident Management is typically empty in this setup and incidents have no occurrence counts, so mentions of incidents or errors are interpreted as Error Tracking issues.

Why does the agent stop before writing a fix?▼

The Skill enforces a hard confirmation gate after root-cause analysis: the agent presents the issues found, the chosen target, the hypothesis with file and commit evidence, and the proposed fix, then waits for approval. Only explicit pre-authorization like "just fix it and PR" bypasses this gate.

Which services and environments are in scope for fixing errors?▼

Only production errors (env:production) in v2 services owned by the repository, such as api, ingest, web, workers, and workflows. Legacy latitude-llm-* v1 services and staging issues are out of scope unless the user explicitly directs otherwise.

Related Skills

Explore More Software Engineering Skills→
githubnextgithubnext
4.9k

github-mcp-server

Configures GitHub MCP server with Docker or hosted modes and selectable tool sets.

Official
Intermediate
GenEducationGenEducation

land-and-deploy

Merge a pull request, wait for CI and deployment, and verify production health with canary checks.

Official
Advanced
finlaysonstudiofinlaysonstudio
2

test

Standardize Vitest testing workflows across JavaScript monorepo packages.

Official
Intermediate
langfuselangfuse
34.0k

debug-issue-with-datadog

Diagnose production issues by correlating Datadog telemetry with Langfuse source code.

Official
Advanced