Automatically fetches your latest Chess.com games, analyzes them with an LLM, and emails you a coaching report with turning points and a training plan.
The AI Agent automates the end-to-end chess game review process: fetch data, analyze it, and deliver a coach-like report by email.
Automates the end-to-end chess game review and coaching delivery.
Fetches your latest Chess.com games from the public API.
Identifies color, result, opponent, rating, and ECO opening code.
Sends the PGN and context to the LLM with a coaching prompt.
Generates a structured HTML coaching report.
Emails the report to your Gmail address.
Allows swapping LLMs (Gemini, GPT-4o, Claude) via simple config.
Automates the end-to-end workflow from game retrieval to coaching delivery. Converts raw game data into concrete, study-ready guidance.
Three-step system to go from game data to a coachable email.
Pulls your current month's Chess.com games from the public API without needing an API key.
Enriches data with color, result, opponent, ratings, and ECO code, then sends PGN and context to the LLM with a coaching prompt.
Generates a clean HTML email and delivers it to your Gmail inbox.
One realistic scenario showing timing and outcomes.
Scenario: A daily run at 7:00 AM fetches the previous day’s games from Chess.com, analyzes a 1-game session, and emails a coachable report by 7:03 AM. The report highlights a critical turning point, lists 3–5 concrete lessons, and includes a 5-minute drill plan. You then use those drills in your morning study and see measurable improvement over the week.
Six roles that gain practical, concrete outcomes.
Wants quick, digestible feedback without manual analysis.
Needs systematic game reviews to build routines.
Uses automated reports to scale coaching and track progress.
Gets clear guidance to support practice at home.
Prepares faster with data-driven study plans.
Provides consistent reviews for a class or cohort.
Connects to Chess.com data sources, LLMs, and email delivery.
Fetches the latest games for analysis.
Delivers the coaching report to your inbox.
Automates the AI agent steps and data flow.
Used as the built-in LLM option for analysis.
Alternative LLM for higher-quality coaching.
Alternative LLM with coaching-friendly tone.
Privacy-focused option via self-hosted LLMs.
Common scenarios where the AI agent adds value.
Practical concerns and setup details.
The AI agent uses Chess.com's public API to fetch a player's most recent games for the current month. It does not require a paid API key for basic access, but rate limits apply for higher-volume usage. You should ensure you are compliant with Chess.com terms of service. If you plan to scale, consider a proper API arrangement. The agent handles batching to respect limits and avoid throttling.
Data safety depends on the LLM and your hosting choice. If you use a cloud-based LLM, review data retention policies and trust settings. For privacy, consider self-hosted LLMs and encrypted connections. You control which reports are emailed and can revoke access at any time. The agent itself does not publish or share your games without your explicit action.
Yes. The AI agent is designed to run with n8n and can operate with self-hosted LLMs. You will need sufficient compute resources for the chosen model. Gemini is optional; you can switch to OpenAI, Claude, or a local LLM. Ensure you meet hardware requirements for the LLM you select. The setup steps cover credential management and hosting considerations.
Multiple models are supported, including Google Gemini, OpenAI GPT-4o, Anthropic Claude, and self-hosted options. In the workflow, you can switch LLMs with a single configuration change and test results side-by-side. The coaching prompts remain consistent, so you can compare quality across models easily. Validate outputs against your expectations and adjust prompts if needed. Model choice affects the depth and tone of coaching.
Yes. The coaching prompts can be adjusted to emphasize specific aspects like endgame technique or time management. You can add or modify sections in the report, such as adding a new drill or tailoring feedback to your preferred study approach. Changes apply to all future reports for consistency. If you have multiple players, you can tailor prompts per player profile. The system supports prompt-level tuning without changing the workflow logic.
Typically a few minutes from fetch to delivery, depending on your model and network latency. The process runs asynchronously in n8n so you can schedule or trigger it on demand. The report is delivered as HTML email, ensuring readability on mobile and desktop. If there are multiple games per day, the agent can batch and include summaries for each game. You can configure the frequency to daily or per-game delivery.
Yes. The workflow can be extended with additional nodes to send reports via Telegram, Slack, Notion, Google Drive, or Discord. You can also archive finished reports to a database or cloud storage for long-term tracking. This extension allows you to centralize coaching material across teams or study groups. Always ensure you have the recipients' consent when sharing coaching content.
Automatically fetches your latest Chess.com games, analyzes them with an LLM, and emails you a coaching report with turning points and a training plan.
