posthog-desktop

A set of working instructions for PostHog’s Electron desktop app, a desktop program built with web technologies. The app lives in its own nested workspace with separate dependencies, tools, and tests.

In plain words
What is it for?
Use it when working under products/desktop, including the Electron app, its agent framework, or related desktop packages and tests.
Why use it?
It prevents desktop work from being mixed up with PostHog’s main web, mobile, or Django codebases, where different rules apply.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/posthog/posthog-foss/posthog-desktop
Any agent
npx skills add PostHog/posthog-foss --skill posthog-desktop
Clone the repo
git clone --depth 1 https://github.com/PostHog/posthog-foss

Made for: Claude Code, Codex.

Per session 179 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,541 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00179 $0.01541
Opus 5 $0.00089 $0.00771
Sonnet 5 $0.00036 $0.00308
Haiku 4.5 $0.00018 $0.00154

Measured today against content hash 7f1a3af54dbd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

posthog-desktop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/posthog-desktop/SKILL.md · 101 lines

How it starts

The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Working in products/desktop

products/desktop/ is the PostHog desktop app (Electron + agent framework), imported from the now-archived PostHog/code repo and kept as a nested standalone workspace: own pnpm-workspace.yaml, own lockfile, own Biome config, Node 22. It is excluded from the root pnpm-workspace.yaml, from ruff/mypy, from root Jest, from oxlint/oxfmt, from stylelint and from pytest. posthog/posthog is the only source of truth: PRs, CI and publishing all happen here, never in PostHog/code.

Read products/desktop/AGENTS.md before writing code in this tree. It is the source of truth for architecture, layer boundaries, DI, and style. This skill only covers the scoping contract around it.

Scope contract

Default: everything you read, edit, run and test lives under products/desktop/. Run all commands with that as cwd. Do not go fishing in frontend/, posthog/, or other products/* for patterns — the desktop tree has its own, and copying monorepo idioms in is a defect.

Step outside only when one of these is true:

  1. The user explicitly tells you to.
  2. A backend API the app calls needs to change. The client is packages/api-client/; the endpoints it hits live in the Django monorepo (see below).
  3. Desktop CI / config that lives at the root by design. The list is fixed: .github/workflows/desktop-*.yml, .github/actions/desktop-*/, .github/scripts/desktop/, and the exclusion entries in pnpm-workspace.yaml, pyproject.toml, pytest.ini, package.json (lint:css), frontend/jest.config.ts, .dockerignore, .oxlintrc.json, .oxfmtrc.json, .config/.markdownlint-cli2.jsonc, .github/workflows/ci-{frontend,storybook,backend}.yml (+ .depot/workflows/ci-backend.yml). Touch these only to keep an exclusion correct; say so when you do.

Anything else outside the tree: stop and ask.

Toolchain — root CLAUDE.md does not apply here

Root monorepo says In products/desktop use
ruff / mypy / pytest nothing — no Python app code here
oxlint + Oxfmt, pnpm --filter=@posthog/frontend fix pnpm lint (Biome check --write), pnpm format
Jest, hogli test Vitest (pnpm test), Playwright (pnpm test:e2e)
pnpm --filter=@posthog/frontend typescript:check pnpm typecheck (turbo, all packages)
Kea logics, typegen:write Zustand stores + InversifyJS services; no kea, no typegen
LemonUI in frontend/ and products/*/frontend/ @posthog/quill + Tailwind + Radix Themes
hogli build:openapi → generated frontend types packages/api-client/src/generated.ts (hand-maintained here)
4-space / repo Prettier habits Biome, 2-space, double quotes

Read the full file on GitHub · 101 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today Changed · +3 lines · +25 tokens per session 7f1a3af54dbd
  2. yesterday First seen · 98 lines · 154 tokens per session scan A 4757bc3b3467

Subscribe to this mod's changes

posthog-desktop is a skill published in the GitHub repository PostHog/posthog-foss (713 stars, last pushed today), licensed MIT. It adds 179 tokens to every session and 1,541 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens