Chess Postmortem Skill
github.com/brumar/chess-postmortem-skills- Category
- AI Agents
- Rank
- No. 1372Tools index
Previous survey · No. 1364 ·
- Type
- AGENT
- Use case
- Research & Education
- Interfaces
- Agent Skill / Plugin
- GitHub
- 76 stars
- Date
About
This Claude Code skill analyzes a player's chess games by combining Claude's vision on board images (instead of PGN notation) with Stockfish engine analysis, then turns live audio or text notes and a vague instruction like 'analyze my last lichess game' into a commented video review. It also reflects on the player's own in-game thinking, which the author found useful for teaching, though a single analysis can take about an hour and cost roughly $15 in API usage.
At a glance
Chess Postmortem turns a played game into a narrative review that uses Stockfish to check the analysis. It is worth reading if a list of engine evaluations leaves you unsure why a move mattered.
Good fit
Chess players who want to revisit their own decisions and can start with a public or otherwise shareable game.
Consider something else when
Use a normal engine review when you only want a quick check of the largest mistakes.
Start here
Pick a short public game and one position you found difficult. Compare the explanation with Stockfish’s alternatives before generating a full narrated review.
Cost and requirements
Claude Code skills with a Stockfish-based workflow. The author’s example uses a game record and locally transcribed audio.
Limits to know
Engine checks ground the move analysis. The explanation and timing can still be wrong, especially when matching spoken thoughts to a move from clock data.
Sources checked 2026-11-05. Source-based assessment.
Why it made the leaderboard
Turns a vague voice note plus a game link into a personalized narrated chess analysis, a novel teaching use of coding-agent tool orchestration beyond code.
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.