![\"\"](\"https:\/\/nodemaven.com\/wp-content\/uploads\/2026\/02\/icon-4.svg\"){kind=icon}
<\/div>\n \n Power your OpenCode workflows with clean residential and mobile proxies. Try NodeMaven from $3.50 with 750 MB included<\/strong><\/p>\n<\/div>\n <\/div>\n \n An OpenCode skill is a folder containing a SKILL.md<\/strong> file. That’s the whole spec.<\/p>\n\n\n\n The file opens with YAML frontmatter (name<\/strong> and description<\/strong> are required) followed by markdown instructions the agent reads when the skill activates.<\/p>\n\n\n\n OpenCode exposes a native skill<\/strong> tool to the model. At session start, it scans every installed skill and reads only the name and description, roughly 100 tokens each. When you give it a task, it checks the list and calls skill({ name: “…” })<\/strong> on anything that matches. The full instructions only load then.<\/p>\n\n\n\n This progressive loading is the whole trick. You can have dozens of skills installed and never pay a token cost for the ones that don’t apply.<\/p>\n\n\n\n OpenCode looks for skills in six locations:<\/strong><\/p>\n\n\n\n Global path is ~\/.config\/opencode\/skills\/<\/strong>, not ~\/.opencode\/skills\/<\/strong>.<\/p>\n\n\n\n This is also why a skill built for Claude Code just works in OpenCode. Same SKILL.md<\/strong> format, same open standard, no conversion needed.<\/p>\n\n\n\n Not every skill on GitHub is worth your context budget.<\/p>\n\n\n\n These made the list based on:<\/p>\n\n\n\n If you’re new to OpenCode skills, don’t install all seven at once. Here’s a quick decision guide:<\/p>\n\n\n\n Instead of ranking these 1 through 7, here’s what actually matters: which job each skill does best.<\/p>\n\n\n\n What it does:<\/strong> Interviews you relentlessly about every branch of a plan before OpenCode writes a single line of code. It walks the decision tree, resolves dependencies one at a time, and gives a recommended answer for each question so you’re reacting instead of stuck.<\/p>\n\n\n\n Best for:<\/strong> Feature specs, architecture decisions, and any plan that feels more solid than it actually is.<\/p>\n\n\n\n Pros:<\/strong><\/p>\n\n\n\n Cons:<\/strong><\/p>\n\n\n\n Example prompt:<\/strong><\/p>\n\n\nWhat are OpenCode skills?<\/h2>\n\n\n\n
Scope<\/strong><\/td> Path<\/strong><\/td> Convention<\/strong><\/td><\/tr><\/thead> Project<\/td> .opencode\/skills\/<\/strong><\/td> OpenCode-native<\/td><\/tr> Project<\/td> .claude\/skills\/<\/strong><\/td> Claude Code-compatible<\/td><\/tr> Project<\/td> .agents\/skills\/<\/strong><\/td> Agent Skills open standard<\/td><\/tr> Global<\/td> ~\/.config\/opencode\/skills\/<\/strong><\/td> OpenCode-native user directory<\/td><\/tr> Global<\/td> ~\/.claude\/skills\/<\/strong><\/td> Claude Code-compatible<\/td><\/tr> Global<\/td> ~\/.agents\/skills\/<\/strong><\/td> Agent Skills open standard<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n How we chose these skills<\/h2>\n\n\n\n
\n
Which skill should you install first?<\/h2>\n\n\n\n
Your situation<\/strong><\/td> Install this first<\/strong><\/td><\/tr><\/thead> You keep correcting wrong assumptions mid-build<\/td> Grill Me<\/td><\/tr> Your sessions die from context limits constantly<\/td> Handoff<\/td><\/tr> OpenCode can’t see current docs or live pages<\/td> Firecrawl<\/td><\/tr> You’re dropped into an unfamiliar codebase<\/td> Understand Anything<\/td><\/tr> Your generated docs read like AI wrote them (they did)<\/td> Stop Slop<\/td><\/tr> Your API bill is higher than it should be<\/td> Caveman<\/td><\/tr> You want a full planning-to-merge pipeline<\/td> Obra Superpowers<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Best OpenCode skills by category<\/h2>\n\n\n\n
1. Best planning skill: Grill Me<\/h3>\n\n\n\n
\n
\n