Guide · Agent skill
How to use the Product Analytics Planner agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The analytics-product skill gives an AI coding agent a structured method for product analytics: naming events, designing funnels and retention cohorts, and sketching a north star metric and dashboard. It is aimed at product managers, developers adding event tracking with PostHog, and analysts who need clear metric definitions. After this guide you will know how to install the skill, trigger it with plain requests, and check what it needs before running any of its example code.
Product Analytics Plannersickn33/agentic-awesome-skills · MIT Download the skill Quick start
- 1Download the ZIP. Download the skill ZIP from the download link on this page.
- 2Unzip into a skills folder. Unzip the folder into ~/.claude/skills/ for personal use, or into .claude/skills/ inside a project.
- 3Or upload it in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.
- 4Check the requirements. Install Python with the posthog and pandas packages if you plan to run the example code, and set POSTHOG_API_KEY only if you need the tracking example.
- 5Describe your analytics task. Ask the agent in plain language, for example to define an activation event or calculate weekly retention.
- 6Review the output before using it. Check event names, windows, denominators and any example numbers, which are illustrative only, before applying them to real data.
What an agent skill is and what this one does
An agent skill is a folder with a SKILL.md file that tells an AI agent how to handle a certain kind of task. When the skill is installed and your request matches its description, the agent follows its method and examples.
This skill covers product analytics: PostHog and Mixpanel, events, funnels, cohorts, retention, DAU/MAU, feature flags, A/B testing, north star metrics, OKRs and product dashboards. It asks the agent to name events in an object_past-verb style, to state the denominator and time window for funnel and retention figures, and to record the product decision, data source, consent rules, time zone and analysis unit before instrumenting anything.
- Event names follow the pattern [object]_[past_verb], for example user_signed_up or conversation_started, rather than vague names like signup or click.
- Funnels and retention cohorts are defined with explicit windows and denominators.
- It includes Python examples for sending events to PostHog and for calculating weekly cohort retention and a north star count from a database.
- Its funnel rates and retention bands are labelled as illustrative and hypothetical, not market benchmarks.
Install the skill from the ZIP download
Download the ZIP of the skill from the download link on this page. Then choose one of two install locations, depending on whether you want the skill for yourself or for one project.
For a personal install, unzip the folder into ~/.claude/skills/. For a project-only install, unzip it into the .claude/skills/ folder inside your project. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/ (available across your projects).
- Single project: .claude/skills/ inside the project folder.
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP.
Trigger the skill with natural requests
You do not need a special command. Describe the analytics task in ordinary language and the agent will use the skill when the request fits its scope. The skill description lists the topics it covers, so mention those terms where you can.
The skill is written in Portuguese with English mixed in, but requests in English still match its topics. Be specific about the product, the user action and the decision you need to make.
- Define an activation event for my onboarding flow and name it using the object_past-verb style.
- Calculate weekly retention cohorts from an events table with user_id, event_date and event_name columns.
- Propose a north star metric for a conversational assistant and a dashboard sketch for it.
- Help me design a signup-to-paid funnel and tell me which windows and denominators to use.
- Set up PostHog event tracking for conversation_started with intent and device properties.
Example: event naming and PostHog tracking
The skill's tracking example uses the PostHog Python library. It reads the project key from the POSTHOG_API_KEY environment variable and sends events with a distinct_id. The host defaults to https://app.posthog.com unless you set POSTHOG_HOST.
Only send events and identify users after you have the authorisation and consent the project requires. The skill itself notes that consent is mentioned only briefly and gives no legal guidance.
- POSTHOG_API_KEY is needed only for the tracking example.
- POSTHOG_HOST is optional and falls back to the default host shown above.
track("user_123", "conversation_started", {"intent": "business_advice", "device": "alexa", "user_tier": "pro"})Example: weekly cohort retention and north star count
The cohort example takes a pandas DataFrame with user_id, event_date and event_name columns and returns a matrix of retention by cohort week and week number. It counts session_started events, so your data needs that event name or the function must be adapted.
The north star example runs a SQL query against a conversations table and counts users with at least three conversations of two minutes or more in a window. The skill asks you to set explicit window limits, use one time zone, and return an aggregate count rather than one row per user. Adapt db.query and calculate_wow_growth to your project before use.
- Requirements for the cohort code: Python with pandas.
- Requirements for the north star query: a database with a conversations table that has created_at, user_id and duration_seconds columns.
- The retention bands and funnel rates in the skill are hypothetical. Replace them with observed baselines from comparable, mature cohorts.
Safety notes: network use and scripts
The skill uses the network when its PostHog examples send events or read feature flags. It does not run scripts on its own; the code it provides is example code that you or the agent run in your environment, so review it before running it against real data or live services.
The skill's feature-flag section says to use the API version you have installed, since the method signatures differ between SDK versions. Check the official PostHog Python documentation linked in the skill before relying on a specific call.
- Uses network: yes, for PostHog calls.
- Runs scripts: no, according to the skill's listed details.
- Requires a PostHog project API key only if you use the tracking example.
- Do not send events or identify users on personal data before consent and authorisation are settled.
What people use it for
- A product manager defines an activation event and a funnel from signup to first meaningful action.
- A developer names and instruments events in a PostHog-based app using the object_past-verb pattern.
- An analyst builds a weekly cohort retention table from an events export.
- A team drafts a north star metric and a simple dashboard query for a conversational product.
Frequently asked questions
Is the analytics-product skill free to use?
The skill is released under the MIT licence, as listed on this page. Check the licence file in the download for the full terms before reusing it in your own project.
What do I need to run the analytics-product skill?
The skill needs Python with the posthog package and pandas for the example code. A PostHog project API key in POSTHOG_API_KEY is needed only for the tracking example, and a database with a conversations table is needed for the north star query.
How do I install an agent skill like this one?
Download the ZIP and unzip it into ~/.claude/skills/ for personal use or into .claude/skills/ in a project. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
Does the skill run scripts on my machine?
According to its listed details, the skill does not run scripts. It provides example Python code that you or the agent run yourself, and it uses the network when PostHog calls are made.
Are the retention and funnel numbers in the skill real benchmarks?
No. The skill labels its funnel rates and retention bands as hypothetical and unsourced. Use them only as a format example and replace them with baselines from your own comparable cohorts.
Does this skill work with Mixpanel too?
The skill's description and topics name Mixpanel, but the code examples it includes are for PostHog. Check the skill text for how much Mixpanel-specific guidance it gives before relying on it for that tool.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you want an agent to apply a clear, consistent method to event naming, funnels, cohorts and north star metrics, especially with PostHog. Pick another resource if you need real industry benchmarks, legal guidance on consent for personal data, or detailed Mixpanel implementation help.
Download the skill