posthog-errors

Generates a weekly cross-project error digest from PostHog error-tracking data via MCP.

714|255|Updated Mar 4, 2026
One-click install
npx skills add https://github.com/aaronjmars/aeon --skill posthog-errors
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: posthog-errors
Source: https://github.com/aaronjmars/aeon/tree/main/skills/posthog-errors
Command: npx skills add https://github.com/aaronjmars/aeon --skill posthog-errors

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Teams running multiple PostHog projects lack a single consolidated view of production errors, forcing them to check each project's error-tracking page manually and making it easy to miss new or spiking issues.

Core Features & Use Cases

  • Cross-project error sweep: Enumerates every project the OAuth grant covers, pulls the last 7 days of error-tracking issues per project, and ranks them by occurrences and users affected.
  • New vs ongoing classification: Flags issues as new this week, spiking, ongoing, or resolved, with an optional week-over-week diff against a stored snapshot.
  • Digest notification with full report: Sends one ranked digest (top issues, per-project table, clickable PostHog links) that links to a committed full report of every issue.
  • Use Case: An engineering lead schedules this weekly to receive one Telegram message summarizing errors across all six of their PostHog projects, with severity set automatically and links to investigate the worst issues.

Quick Start

Run the posthog-errors skill to send me a digest of all errors across my PostHog projects from the last 7 days.

Frequently Asked Questions about posthog-errors

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

FAQPage Schema
How do I get a weekly error digest from PostHog?

Connect PostHog via the dashboard MCP panel's OAuth flow, then run this skill on a weekly schedule. It enumerates all projects in the grant, pulls 7 days of error-tracking issues per project, ranks them by impact, and sends one digest notification.

How does the skill rank PostHog errors by impact?

Impact ranking combines occurrence counts and users affected, with a 50-user error typically outranking a noisy 5-user one at similar volume. Issues are also classified as new this week, spiking, ongoing, or resolved based on first-seen and last-seen timestamps.

Can I limit the digest to specific PostHog projects or a different time window?

Yes, the optional var parameter accepts a comma-list of project name substrings to filter scope and a window like 14d or 30d to change the default 7-day range. For example, 'web, api, 14d' limits the sweep to matching projects over two weeks.

Why are some of my PostHog projects missing from the digest?

Missing projects mean the OAuth grant does not cover their organizations, since scope is bounded by what was authorized at Connect. Re-connect PostHog in the dashboard and authorize the missing organizations to widen coverage.

What happens when the PostHog MCP server is not connected or the token expired?

If no PostHog MCP tools are callable, the skill logs POSTHOG_NOT_CONNECTED and notifies once to reconnect in the dashboard. On 401 or invalid token errors it logs POSTHOG_AUTH_STALE, asks for a one-time re-connect, and exits with clearly marked partial results.

Does the skill modify or resolve errors in PostHog?

No, the skill runs in read-only mode and never resolves, mutes, or edits anything in PostHog. It only reports; a human decides what to act on via the linked PostHog issue pages.