[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-1409":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":14,"forks30d":14,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":35,"readmeContent":36,"aiSummary":37,"trendingCount":14,"starSnapshotCount":14,"syncStatus":38,"lastSyncTime":39,"discoverSource":40},1409,"opencode-power-pack","waybarrios\u002Fopencode-power-pack","waybarrios","Eleven Claude Code skills ported to OpenCode: code-review, security-review, feature-dev, frontend-design + 7 more. One config line, one plugin.",null,"JavaScript",381,34,1,0,7,14,53,21,4.63,"MIT License",false,"main",true,[25,26,27,28,29,30,31,32,33,34],"ai-agents","anthropic","claude-code","code-review","developer-tools","llm-tools","opencode","plugin","security-audit","skills","2026-06-12 02:00:27","\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Flogo.svg\" alt=\"opencode-power-pack\" width=\"100%\" \u002F>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ci>Eleven Claude Code skills, ported to OpenCode.\u003Cbr\u002F>\n  Code review, security audit, feature dev, frontend design, and the rest of the kit — installable in one line.\u003C\u002Fi>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack\u002Fblob\u002Fmain\u002FLICENSE\">\u003Cimg alt=\"License: MIT\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-brightgreen?style=flat-square\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack\u002Fstargazers\">\u003Cimg alt=\"GitHub stars\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fwaybarrios\u002Fopencode-power-pack?style=flat-square&color=FFD60A\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack\u002Fcommits\u002Fmain\">\u003Cimg alt=\"Last commit\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002Fwaybarrios\u002Fopencode-power-pack?style=flat-square\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack\u002Fissues\">\u003Cimg alt=\"Issues\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002Fwaybarrios\u002Fopencode-power-pack?style=flat-square\">\u003C\u002Fa>\n  \u003Cimg alt=\"Skills\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fskills-11-FFD60A?style=flat-square&labelColor=0B0F14\">\n  \u003Cimg alt=\"OpenCode\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fopencode-compatible-0B0F14?style=flat-square&labelColor=FFD60A\">\n  \u003Cimg alt=\"Claude Code compat\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fclaude--code-skills_format-D97706?style=flat-square\">\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"#installation\">\u003Cb>Install\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"#whats-inside\">\u003Cb>Skills\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"#slash-commands\">\u003Cb>Commands\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"#how-it-works\">\u003Cb>How it works\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"#acknowledgments\">\u003Cb>Credits\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"LICENSE\">\u003Cb>License\u003C\u002Fb>\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ci>Built on top of \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code\">anthropics\u002Fclaude-code\u003C\u002Fa>,\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills\">anthropics\u002Fskills\u003C\u002Fa>,\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code-security-review\">anthropics\u002Fclaude-code-security-review\u003C\u002Fa>,\n  and \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fobra\u002Fsuperpowers\">obra\u002Fsuperpowers\u003C\u002Fa>. See \u003Ca href=\"#acknowledgments\">Acknowledgments\u003C\u002Fa>.\u003C\u002Fi>\n\u003C\u002Fp>\n\n---\n\n## Why this exists\n\nOpenCode reads Claude Code's `SKILL.md` format natively, but **most of Anthropic's official Claude Code plugins put their value in `commands\u002F` and `agents\u002F`** — and those are Claude-Code-only. So if you want `\u002Fcode-review`, `\u002Fsecurity-review`, or `\u002Ffeature-dev` in OpenCode, copy-paste won't get you there.\n\nThis pack does the **translation**: the multi-agent workflows from those plugins are rewritten as OpenCode-compatible skills, so the methodology survives the platform jump. Plus a few direct ports of skills that already lived in Anthropic's `skills\u002F` repo.\n\nIt pairs nicely with **[obra\u002Fsuperpowers](https:\u002F\u002Fgithub.com\u002Fobra\u002Fsuperpowers)**, which provides the meta-workflow skills (brainstorming, TDD, executing-plans). This pack adds the domain-specific muscle.\n\n---\n\n## What's inside\n\n\u003Ctable>\n\u003Ctr>\n\u003Cth align=\"left\">Category\u003C\u002Fth>\n\u003Cth align=\"left\">Skill\u003C\u002Fth>\n\u003Cth align=\"left\">Source\u003C\u002Fth>\n\u003Cth align=\"left\">Purpose\u003C\u002Fth>\n\u003C\u002Ftr>\n\n\u003Ctr>\u003Ctd rowspan=\"2\">\u003Cb>Review\u003C\u002Fb>\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>code-review\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · plugins\u002Fcode-review\u003C\u002Ftd>\n\u003Ctd>Multi-agent PR review with confidence-filtered cross-checks and reproduction scenarios\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>security-review\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · claude-code-security-review\u003C\u002Ftd>\n\u003Ctd>OWASP-bucketed, three-stage filtering, requires concrete attack PoC per finding\u003C\u002Ftd>\u003C\u002Ftr>\n\n\u003Ctr>\u003Ctd rowspan=\"4\">\u003Cb>Feature dev\u003C\u002Fb>\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>feature-dev\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · plugins\u002Ffeature-dev\u003C\u002Ftd>\n\u003Ctd>Seven-phase guided workflow: discovery → exploration → questions → architecture → impl → review → summary\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>code-explorer\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · feature-dev\u002Fagents\u003C\u002Ftd>\n\u003Ctd>Deep codebase analysis sub-task — traces a feature end-to-end\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>code-architect\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · feature-dev\u002Fagents\u003C\u002Ftd>\n\u003Ctd>Decisive architecture blueprint sub-task with file-level implementation map\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>code-reviewer\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · feature-dev\u002Fagents\u003C\u002Ftd>\n\u003Ctd>Two-pass adversarial review sub-task with explicit edge-case checklist\u003C\u002Ftd>\u003C\u002Ftr>\n\n\u003Ctr>\u003Ctd>\u003Cb>Design\u003C\u002Fb>\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>frontend-design\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>copied · plugins\u002Ffrontend-design\u003C\u002Ftd>\n\u003Ctd>Distinctive, production-grade UI generation that avoids generic AI aesthetics\u003C\u002Ftd>\u003C\u002Ftr>\n\n\u003Ctr>\u003Ctd rowspan=\"2\">\u003Cb>Authoring\u003C\u002Fb>\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>mcp-builder\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>copied · skills\u002Fmcp-builder\u003C\u002Ftd>\n\u003Ctd>Build high-quality MCP servers (Python or TypeScript)\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>skill-creator\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>adapted · skills\u002Fskill-creator\u003C\u002Ftd>\n\u003Ctd>Author new SKILL.md files with progressive-disclosure structure\u003C\u002Ftd>\u003C\u002Ftr>\n\n\u003Ctr>\u003Ctd rowspan=\"2\">\u003Cb>Project memory\u003C\u002Fb>\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>agents-md-improver\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>adapted · plugins\u002Fclaude-md-management\u003C\u002Ftd>\n\u003Ctd>Audit and update \u003Ccode>AGENTS.md\u003C\u002Fcode> \u002F \u003Ccode>CLAUDE.md\u003C\u002Fcode> against the current codebase\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Ccode>agents-md-revise\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>translated · claude-md-management\u002Fcommands\u003C\u002Ftd>\n\u003Ctd>Capture session learnings into project rules — complement to \u003Ccode>improver\u003C\u002Fcode>\u003C\u002Ftd>\u003C\u002Ftr>\n\n\u003C\u002Ftable>\n\n---\n\n## Installation\n\n### Prerequisites\n\n- **OpenCode** installed and on your PATH. If you do not have it: \u003Chttps:\u002F\u002Fopencode.ai>\n- **git** (used by OpenCode to fetch the plugin)\n- **Bun** is bundled with OpenCode, so the plugin's `npm install` step at startup is automatic\n\n### Pick an install method\n\nTwo paths that work today; pick whichever fits how you plan to use the pack.\n\n| | **A — Install from GitHub** | **B — Install from a local git clone** |\n|---|---|---|\n| Best for | Just using the pack | Tinkering, contributing, or running offline |\n| Requires | Network during first install | A local clone of the repo |\n| Updates | `git pull` + cache clear (the cache holds Bun's checkout) | `git pull` updates everything live; symlinks pick up changes immediately |\n| Internet at runtime | Only at first install \u002F on cache miss | Never |\n\nBoth methods share **steps 2 (symlink commands) and 3 (restart)**. Only step 1 differs.\n\n---\n\n### A — Install from GitHub\n\n**Step 1A.** Add to `~\u002F.config\u002Fopencode\u002Fopencode.json`:\n\n```jsonc\n{\n  \"$schema\": \"https:\u002F\u002Fopencode.ai\u002Fconfig.json\",\n  \"plugin\": [\n    \"opencode-power-pack@git+https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack.git\"\n  ]\n}\n```\n\nIf you already use other plugins (e.g. `superpowers`), keep all of them in the array:\n\n```jsonc\n{\n  \"plugin\": [\n    \"superpowers@git+https:\u002F\u002Fgithub.com\u002Fobra\u002Fsuperpowers.git\",\n    \"opencode-power-pack@git+https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack.git\"\n  ]\n}\n```\n\nTo pin a specific tag (recommended once releases exist):\n\n```jsonc\n\"opencode-power-pack@git+https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack.git#v0.1.0\"\n```\n\nYou still need a local copy of the repo for **step 2** (the slash command files live there). Clone it next to wherever you keep code:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack.git ~\u002Fcode\u002Fopencode-power-pack\n```\n\nThen jump to **step 2 (symlink commands)** below.\n\n---\n\n### B — Install from a local git clone\n\nThis avoids the GitHub round-trip entirely. The plugin lives on your disk, and OpenCode reads it via a `file:\u002F\u002F` URL.\n\n**Step 1B.1.** Clone the repo wherever you keep code:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fwaybarrios\u002Fopencode-power-pack.git ~\u002Fcode\u002Fopencode-power-pack\n```\n\nIf the repo is not yet on GitHub and you only have it locally, skip the clone — just point at the existing directory.\n\n**Step 1B.2.** Add to `~\u002F.config\u002Fopencode\u002Fopencode.json`, using the **absolute path** to your clone:\n\n```jsonc\n{\n  \"$schema\": \"https:\u002F\u002Fopencode.ai\u002Fconfig.json\",\n  \"plugin\": [\n    \"opencode-power-pack@git+file:\u002F\u002F\u002FUsers\u002Fyou\u002Fcode\u002Fopencode-power-pack\"\n  ]\n}\n```\n\nNotes:\n\n- The `file:\u002F\u002F` URL needs to be an **absolute path** with three slashes (`file:\u002F\u002F\u002FUsers\u002F...`) and no trailing slash.\n- The directory must be a git repo (have a `.git\u002F`). If you copied the files without git, run `git init && git add . && git commit -m init` inside the directory first — Bun's git-style installer requires a real git tree.\n- Updating is just `cd ~\u002Fcode\u002Fopencode-power-pack && git pull`, then clear the cache and restart (see [Updating](#updating)).\n\nThen continue to **step 2** below.\n\n### 2. Symlink the slash commands (both methods)\n\nThe plugin auto-registers the skills directory programmatically. Slash commands, however, need to live in OpenCode's canonical commands path; the plugin ships physical markdown files under `commands\u002F` for you to symlink in.\n\n```bash\nmkdir -p ~\u002F.config\u002Fopencode\u002Fcommands\nln -s ~\u002Fcode\u002Fopencode-power-pack\u002Fcommands\u002F*.md ~\u002F.config\u002Fopencode\u002Fcommands\u002F\n```\n\nAdjust the source path if you cloned somewhere other than `~\u002Fcode\u002Fopencode-power-pack`. If a command of the same name already exists in `~\u002F.config\u002Fopencode\u002Fcommands\u002F`, that file wins — `ln -s` will refuse to overwrite, which is the desired behavior.\n\nIf you prefer **copies** over symlinks (e.g. you do not want `git pull` to silently change your commands):\n\n```bash\ncp ~\u002Fcode\u002Fopencode-power-pack\u002Fcommands\u002F*.md ~\u002F.config\u002Fopencode\u002Fcommands\u002F\n```\n\nTrade-off: copies are static — you have to re-copy after every `git pull`. Symlinks track the working tree live.\n\n### 3. Restart OpenCode (both methods)\n\n```bash\n# kill ALL opencode processes (not just the active TUI)\npkill -f opencode\n\n# clear the npm-style plugin cache so the new version is fetched fresh\nrm -rf ~\u002F.cache\u002Fopencode\u002Fnode_modules\u002Fopencode-power-pack 2>\u002Fdev\u002Fnull\n\n# start again\nopencode\n```\n\n### 4. Verify\n\nIn a new OpenCode session, run:\n\n```\nList the skills you have available.\n```\n\nYou should see the eleven skills under the `opencode-power-pack:` namespace (or unprefixed, depending on your OpenCode version).\n\nThen hit **`ctrl+p`** to open the command palette and look for any of:\n\n- `code-review`\n- `security-review`\n- `feature-dev`\n- `frontend-design`\n- `agents-md-improver`\n\nIf they show up, you're done. If not, see [Troubleshooting](#troubleshooting).\n\n---\n\n## Updating\n\n```bash\n# 1. Pull the latest skills + commands\ncd ~\u002Fcode\u002Fopencode-power-pack\ngit pull\n\n# 2. If new skills were added, the existing symlinks are unaffected,\n#    but new commands need to be linked in:\nln -s ~\u002Fcode\u002Fopencode-power-pack\u002Fcommands\u002F*.md ~\u002F.config\u002Fopencode\u002Fcommands\u002F 2>\u002Fdev\u002Fnull\n\n# 3. Clear the plugin cache and restart OpenCode\nrm -rf ~\u002F.cache\u002Fopencode\u002Fnode_modules\u002Fopencode-power-pack\npkill -f opencode\nopencode\n```\n\nIf you pinned a version in `opencode.json` (e.g. `#v0.1.0`), bump the tag in the JSON before restarting, otherwise OpenCode keeps using the pinned commit.\n\n---\n\n## Uninstalling\n\n```bash\n# 1. Remove from opencode.json\n#    delete the \"opencode-power-pack@...\" line from the \"plugin\" array\n\n# 2. Remove the symlinked commands\nfor f in ~\u002Fcode\u002Fopencode-power-pack\u002Fcommands\u002F*.md; do\n  rm -f ~\u002F.config\u002Fopencode\u002Fcommands\u002F\"$(basename \"$f\")\"\ndone\n\n# 3. Remove the plugin cache\nrm -rf ~\u002F.cache\u002Fopencode\u002Fnode_modules\u002Fopencode-power-pack\n\n# 4. Restart OpenCode\npkill -f opencode\nopencode\n\n# 5. Optional: delete the local clone\nrm -rf ~\u002Fcode\u002Fopencode-power-pack\n```\n\nTo remove **just one skill** (keeping the rest), delete its symlink only:\n\n```bash\nrm ~\u002F.config\u002Fopencode\u002Fcommands\u002Fcode-review.md\n```\n\nThe skill itself remains discoverable via the native `skill` tool unless you also remove it from the skills directory, but the slash command will no longer appear.\n\n---\n\n## Troubleshooting\n\n| Symptom | Cause | Fix |\n|---|---|---|\n| `\u002Fcode-review` not in command palette | Symlinks not in `~\u002F.config\u002Fopencode\u002Fcommands\u002F` | Re-run the `ln -s` step from Installation |\n| `\u002Fcode-review` shows the old meta-prompt (\"If the skill is not yet loaded, load it via the native skill tool…\") | Stale plugin cache from a pre-`v0.1.0` build | `rm -rf ~\u002F.cache\u002Fopencode\u002Fnode_modules\u002Fopencode-power-pack && pkill -f opencode && opencode` |\n| Skills do not appear when you run \"list available skills\" | Plugin entry missing or misspelled in `opencode.json` | Validate JSON: `python3 -m json.tool ~\u002F.config\u002Fopencode\u002Fopencode.json` |\n| Plugin install fails with a git error | Bad URL, network issue, or private repo | Check the `git+...` URL works manually: `git ls-remote \u003Curl>` |\n| Symlink command fails on macOS with permission denied | `~\u002F.config\u002Fopencode\u002Fcommands\u002F` does not exist | `mkdir -p ~\u002F.config\u002Fopencode\u002Fcommands` first |\n| Command appears but model \"rushes\" and gives a one-line answer | Local model not following multi-step instructions | Use a stronger backing model. The skill content is correct; small models may still skim. The skills mark themselves as \"expected to take multiple minutes\" — bigger models honor that. |\n| Want to disable for one project | Project-level override | Add `\"plugin\": []` (empty) to `\u003Cproject>\u002F.opencode\u002Fopencode.json` |\n\nIf none of the above match: `tail -f ~\u002F.cache\u002Fopencode\u002Flog\u002F*.log` while you start OpenCode and look for plugin-load errors.\n\n---\n\n---\n\n## Slash commands\n\nEach skill is exposed as a slash command. The command body inlines the **full skill workflow**, so the model receives the actual instructions as its prompt — not a meta-instruction telling it to load something else.\n\n```\n\u002Fcode-review                Multi-agent PR review with cross-check and reproduction scenarios\n\u002Fsecurity-review            OWASP-bucketed audit with three-stage filtering and PoC requirement\n\u002Ffeature-dev                Start the 7-phase guided feature workflow\n\u002Fcode-explorer              Trace a feature deeply across the codebase\n\u002Fcode-architect             Produce a complete architecture blueprint for a feature\n\u002Fcode-reviewer              Two-pass adversarial review of a small change set\n\u002Ffrontend-design            Generate a distinctive frontend with a bold aesthetic direction\n\u002Fmcp-builder                Guide MCP server creation\n\u002Fskill-creator              Author a new SKILL.md\n\u002Fagents-md-improver         Audit and improve AGENTS.md \u002F CLAUDE.md\n\u002Fagents-md-revise           Capture session learnings into the rules file\n```\n\n### Examples\n\n```\n\u002Fcode-review --comment\n\u002Fcode-review can you review https:\u002F\u002Fgithub.com\u002Fowner\u002Frepo\u002Fpull\u002F449\n\u002Ffeature-dev add a logout button to the topbar\n\u002Fsecurity-review\n\u002Ffrontend-design pricing page, brutalist tone, single-screen\n\u002Fagents-md-improver\n```\n\nIf you have a command of the same name in `~\u002F.config\u002Fopencode\u002Fcommands\u002Ffoo.md` or under `command` in `opencode.json`, **your version wins** — the plugin only registers the skill discovery path, never overwrites a command.\n\n---\n\n## How it works\n\n```\n┌─────────────────────────── opencode-power-pack ────────────────────────────┐\n│                                                                            │\n│   .opencode\u002Fplugins\u002Fopencode-power-pack.js                                 │\n│   ├── exports OpencodePowerPack(ctx) → { config(c) }                       │\n│   └── pushes  skills\u002F  into  c.skills.paths   (live config singleton)      │\n│                                                                            │\n│   skills\u002F\u003Cname>\u002FSKILL.md                                                   │\n│   └── frontmatter: name + description (+ license)                          │\n│       body: the actual workflow the model executes                         │\n│                                                                            │\n│   commands\u002F\u003Cname>.md          ──symlink──▶  ~\u002F.config\u002Fopencode\u002Fcommands\u002F   │\n│   ├── frontmatter: description (for menu)                                  │\n│   └── body: full inlined SKILL.md content + $ARGUMENTS                     │\n│                                                                            │\n└────────────────────────────────────────────────────────────────────────────┘\n```\n\n**Two surfaces, one source of truth.** The skills directory feeds OpenCode's native `skill` tool (so the model can load any skill on demand). The commands directory feeds the slash-command palette (so you can invoke a skill directly with `\u002F\u003Cname>`). Both are generated from the same `SKILL.md` files — edit the SKILL.md, regenerate the command file with the script in the repo, and both paths stay in sync.\n\nThe plugin entry-point only registers the skills path. Slash commands are **physical files**, not programmatic — this avoids OpenCode's runtime-config caching gotchas and means you can edit, symlink, or selectively delete commands without touching code.\n\n---\n\n## Scope and non-goals\n\n| In scope | Out of scope |\n|---|---|\n| Porting Claude Code skills where the methodology is portable | Claude Code slash commands ported as Claude-Code-style commands |\n| Translating commands and agents into SKILL.md format | Claude Code hooks |\n| Direct copies of `anthropics\u002Fskills` skills with attribution | Claude Code output styles |\n| OpenCode-native slash commands generated from skills | Anything that breaks if you also use Claude Code |\n\n---\n\n## Contributing\n\nPull requests welcome, especially for:\n\n- New skill ports (Trail of Bits security skills, language-specific packs from `wshobson\u002Fagents`, document skills, etc.)\n- Improvements to existing skill instructions based on real-world failure modes\n- Tooling to keep `commands\u002F*.md` in sync with `skills\u002F*\u002FSKILL.md` automatically\n\nSkill format must follow the OpenCode spec:\n\n- Directory name matches the `name` frontmatter field\n- Lowercase alphanumeric with single hyphens, regex `^[a-z0-9]+(-[a-z0-9]+)*$`\n- `description` 1–1024 chars, specific enough to trigger correctly\n- Body in markdown, ideally under 500 lines\n- Cite the upstream source in `license` frontmatter when porting\n\n---\n\n## Acknowledgments\n\n**This package is not original work.** Almost everything in `skills\u002F` is either a direct copy or a translation of skills, commands, or agent definitions written by **Anthropic** for Claude Code, plus one direct adaptation of the OpenCode plugin pattern from **Jesse Vincent (obra)**. The credit belongs to them; this repo's contribution is the *porting and packaging* for OpenCode.\n\n### Upstream sources\n\n| Upstream | Project | What we use from it |\n|---|---|---|\n| Anthropic | [`anthropics\u002Fclaude-code\u002Fplugins\u002Fcode-review`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code\u002Ftree\u002Fmain\u002Fplugins\u002Fcode-review) | The four-reviewer parallel-review methodology that became `code-review` |\n| Anthropic | [`anthropics\u002Fclaude-code\u002Fplugins\u002Ffeature-dev`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code\u002Ftree\u002Fmain\u002Fplugins\u002Ffeature-dev) | The seven-phase workflow and the three sub-agents (`code-explorer`, `code-architect`, `code-reviewer`) |\n| Anthropic | [`anthropics\u002Fclaude-code\u002Fplugins\u002Ffrontend-design`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code\u002Ftree\u002Fmain\u002Fplugins\u002Ffrontend-design) | The `frontend-design` skill (direct copy) |\n| Anthropic | [`anthropics\u002Fclaude-code-security-review`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code-security-review) | The `\u002Fsecurity-review` slash command, translated and extended |\n| Anthropic | [`anthropics\u002Fskills\u002Fskills\u002Fmcp-builder`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills\u002Ftree\u002Fmain\u002Fskills\u002Fmcp-builder) | The `mcp-builder` skill (direct copy, references stripped) |\n| Anthropic | [`anthropics\u002Fskills\u002Fskills\u002Fskill-creator`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills\u002Ftree\u002Fmain\u002Fskills\u002Fskill-creator) | The `skill-creator` skill (adapted, eval-tooling references trimmed) |\n| Anthropic | [`anthropics\u002Fclaude-plugins-official\u002F...\u002Fclaude-md-management`](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-plugins-official\u002Ftree\u002Fmain\u002Fplugins\u002Fclaude-md-management) | Renamed to `agents-md-improver` and `agents-md-revise`; covers AGENTS.md too |\n| Jesse Vincent (obra) | [`obra\u002Fsuperpowers`](https:\u002F\u002Fgithub.com\u002Fobra\u002Fsuperpowers) | The OpenCode plugin pattern (`config.skills.paths.push(...)`) used in `.opencode\u002Fplugins\u002Fopencode-power-pack.js` |\n\n### What this repo actually contributes\n\n- The translation of Claude-Code-only artifacts (commands and agents) into the SKILL.md format OpenCode reads natively\n- Renaming and adapting `claude-md-improver` \u002F `revise-claude-md` to cover both `AGENTS.md` (OpenCode-native) and `CLAUDE.md` (compatibility)\n- Deepening the review skills (`code-review`, `code-reviewer`, `security-review`) with extra reviewers, multi-pass adversarial analysis, mandatory category coverage, and concrete-PoC requirements — the originals were already strong; the ports try to compensate for the smaller models people sometimes run under OpenCode by being more directive\n- Bundling everything as a one-line OpenCode plugin, plus generating physical slash command files that inline each skill's full content\n\nEach individual SKILL.md frontmatter includes a `license` field naming its specific upstream. The wrapper code (the plugin JS, the README, the LICENSE) is the only original-work portion of this repo.\n\nIf you are one of the upstream authors and you'd like the attribution worded differently — or removed — open an issue and we'll fix it.\n\n## License\n\nMIT for the wrapper code and original work in this repo. Each ported skill cites its upstream source in its frontmatter; upstream Anthropic projects are also MIT-licensed. See [LICENSE](LICENSE) for full attribution.\n","opencode-power-pack 项目将Eleven Claude Code的11项技能移植到了OpenCode平台，涵盖代码审查、安全审计、功能开发、前端设计等。该项目使用JavaScript编写，通过简单的配置即可集成这些功能强大的插件。其核心优势在于能够无缝地将Claude Code的多代理工作流程转换为OpenCode兼容的技能集，特别适合需要跨平台迁移AI辅助开发工具的开发者或团队使用。此外，它与obra\u002Fsuperpowers项目的结合使用可以进一步增强开发效率和代码质量。",2,"2026-06-11 02:43:33","CREATED_QUERY"]