Skip to content
Back to search

opencode-power-pack

MarketplaceWorks with Claude CodeGitHub

Skills for Claude Code, Codex, OpenCode 1 and 2, and Pi.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/waybarrios/opencode-power-pack with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/waybarrios/opencode-power-pack'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/waybarrios--opencode-power-pack/badge.svg)](/listings/waybarrios--opencode-power-pack)

HTML

<a href="/listings/waybarrios--opencode-power-pack"><img src="/api/listings/waybarrios--opencode-power-pack/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Prerequisites

- Git - Node.js 20.11.0 or newer, including npm/npx, for [selective installation](#selective-install-with-npm) and the sandbox runner - One supported host: - A current Claude Code installation with plugin support: <https://code.claude.com/docs/en/plugins> - OpenCode 1.18.7 or newer, including OpenCode 2: <https://opencode.ai> - A current Codex CLI or Codex desktop environment with plugin support: <https://developers.openai.com/codex/> - A current Pi coding agent installation: <https://pi.dev/>

Claude Code

Add the repository marketplace and install its plugin from inside Claude Code:

/plugin marketplace add waybarrios/opencode-power-pack
/plugin install opencode-power-pack@opencode-power-pack

Run `/reload-plugins` if prompted. Claude Code loads plugin skills from the repository-level `skills/` directory and prefixes them with the plugin namespace, preventing collisions with built-in or separately installed skills. For local development, run `claude --plugin-dir .` from a clone of this repository.

Codex CLI And Desktop

Add this repository as a marketplace, then install the plugin:

codex plugin marketplace add waybarrios/opencode-power-pack --ref main
codex plugin add opencode-power-pack@opencode-power-pack
codex plugin list --marketplace opencode-power-pack

Start a new Codex session after installation so the fifty-five bundled skills are loaded. Use `/plugins` to inspect the installed plugin or `$` to select one of its skills explicitly. Codex plugin packaging follows the [official plugin structure](https://developers.openai.com/plugins/build/plugins).

For a smaller personal or repository-specific set, use the [selective npm installer](#selective-install-with-npm) instead of the full plugin.

OpenCode From GitHub

On OpenCode 2:

opencode plugin add "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"

On OpenCode 1:

opencode plugin --global "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"

Restart OpenCode after installation. One plugin adapts to the host version: OpenCode 2 registers all fifty-five skills and the three specialist agents through the V2 plugin API, and OpenCode 1 keeps the original `config.skills.paths` and `config.agent` hook. Either way the skills are invocable by name and the three roles are registered as read-only subagents.

The compatibility smoke tests cover OpenCode **1.18.7, 2.0.14, and 2.0.32**. The `SKILL.md` format stays the same across these hosts.

To pin a release, append its tag to the same spec. Choose a release that includes the V2 port when using OpenCode 2; `v0.5.0` predates it:

opencode plugin add "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git#<tag>"

OpenCode From A Local Clone

git clone https://github.com/waybarrios/opencode-power-pack.git ~/code/opencode-power-pack
opencode plugin add "opencode-power-pack@git+file:///home/you/code/opencode-power-pack"

Use an absolute `file://` URL adjusted for your operating system. The target directory must be a Git repository. On OpenCode 1, replace `plugin add` with `plugin --global` as shown above.

Pi From GitHub

pi install git:github.com/waybarrios/opencode-power-pack

Pi installs the Git package and loads the `skills/` directory declared in `package.json`. Use `-l` to save it to project settings instead of user settings. See the [Pi package documentation](https://pi.dev/docs/latest/packages) for source, update, filtering, and trust behavior.

Pi can also narrow an installed package through `pi config`. The npm installer is useful when you want the same selected set to be discovered by Pi, Codex, and OpenCode from `.agents/skills`.

Verify Claude Code

From a local clone, validate both the plugin and marketplace manifests:

claude plugin validate . --strict

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 23 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "$schema": "https://json.schemastore.org/claude-code-marketplace.json",
  3. "name": "opencode-power-pack",
  4. "owner": {
  5. "name": "Wayner Barrios",
  6. "url": "https://github.com/waybarrios"
  7. },
  8. "description": "Skills for Claude Code, Codex, OpenCode 1 and 2, and Pi.",
  9. "plugins": [
  10. {
  11. "name": "opencode-power-pack",
  12. "source": "./",
  13. "description": "Fifty-five workflows for review, security, feature development, frontend work, writing, authoring, and ML tools.",
  14. "category": "productivity",
  15. "tags": [
  16. "software-development",
  17. "code-review",
  18. "security",
  19. "skills"
  20. ]
  21. }
  22. ]
  23. }
README.md · README · 698 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <p align="center">
  2. <img src="assets/logo.svg" alt="opencode-power-pack" width="100%" />
  3. </p>
  4. <h1 align="center">OpenCode Power Pack for Claude Code, Codex, OpenCode 1 + 2, and Pi</h1>
  5. <p align="center">
  6. <i>Fifty-five portable workflows for Claude Code, Codex, OpenCode 1 and 2, and Pi.<br/>
  7. Code review, security audit, feature development, frontend design, project memory, and authoring tools.</i>
  8. </p>
  9. <p align="center">
  10. <a href="THIRD_PARTY_NOTICES.md"><img alt="License: MIT and Apache-2.0" src="https://img.shields.io/badge/license-MIT%20%2B%20Apache--2.0-brightgreen?style=flat-square"></a>
  11. <a href="https://github.com/waybarrios/opencode-power-pack/stargazers"><img alt="GitHub stars" src="https://img.shields.io/github/stars/waybarrios/opencode-power-pack?style=flat-square&color=FFD60A"></a>
  12. <a href="https://github.com/waybarrios/opencode-power-pack/commits/main"><img alt="Last commit" src="https://img.shields.io/github/last-commit/waybarrios/opencode-power-pack?style=flat-square"></a>
  13. <a href="https://github.com/waybarrios/opencode-power-pack/issues"><img alt="GitHub issues" src="https://img.shields.io/github/issues/waybarrios/opencode-power-pack?style=flat-square"></a>
  14. <a href="https://github.com/hashgraph-online/hol-guard"><img alt="HOL Guard scanner" src="https://img.shields.io/badge/HOL%20Guard-scanned-00a67e?style=flat-square"></a>
  15. <img alt="Skills: 55" src="https://img.shields.io/badge/skills-55-FFD60A?style=flat-square&labelColor=0B0F14">
  16. <img alt="Claude Code plugin" src="https://img.shields.io/badge/Claude_Code-plugin-0B0F14?style=flat-square&labelColor=FFD60A">
  17. <img alt="OpenCode 1.18.7+ / 2.x" src="https://img.shields.io/badge/opencode-1.18.7%2B%20%2F%202.x-0B0F14?style=flat-square&labelColor=FFD60A">
  18. <img alt="Codex plugin" src="https://img.shields.io/badge/Codex-plugin-0B0F14?style=flat-square&labelColor=FFD60A">
  19. <img alt="Pi package" src="https://img.shields.io/badge/Pi-package-0B0F14?style=flat-square&labelColor=FFD60A">
  20. </p>
  21. <p align="center">
  22. <a href="#installation"><b>Install</b></a> ·
  23. <a href="https://skills.waybarrios.com/docs/"><b>Documentation</b></a> ·
  24. <a href="#opencode-2-support"><b>OpenCode 2</b></a> ·
  25. <a href="#selective-install-with-npm"><b>Select Skills</b></a> ·
  26. <a href="#claude-code-quick-install"><b>Claude Code</b></a> ·
  27. <a href="#codex-quick-install"><b>Codex Quick Install</b></a> ·
  28. <a href="#pi-quick-install"><b>Pi Quick Install</b></a> ·
  29. <a href="#whats-inside"><b>Skills</b></a> ·
  30. <a href="#sandbox-capability-contract"><b>Sandbox Contract</b></a> ·
  31. <a href="#invocation"><b>Invocation</b></a> ·
  32. <a href="#how-it-works"><b>Architecture</b></a> ·
  33. <a href="#acknowledgments"><b>Credits</b></a>
  34. </p>
  35. ## OpenCode 2 support
  36. Version 0.6.0 supports OpenCode 2 and keeps OpenCode 1.18.7 support. Both hosts load all 55 skills and the three feature-development specialists. Smoke tests cover OpenCode 1.18.7, 2.0.14, and 2.0.32, including effective read-only permissions on V2.
  37. Install the plugin in OpenCode 2:
  38. ```bash
  39. opencode plugin add "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  40. ```
  41. Release `v0.6.0` and npm `0.6.0` include the plugin for OpenCode 1 and 2. Tag `v0.5.0` and npm `0.5.0` have only the OpenCode 1 plugin.
  42. For host commands, update steps, and the current catalog, see the [documentation](https://skills.waybarrios.com/docs/) and [skill reference](https://skills.waybarrios.com/docs/skills/). The [OpenCode compatibility guide](docs/opencode-compatibility.md) covers source selection and V2 permissions.
  43. ## Native sandboxes in v0.5.0
  44. OpenCode Power Pack now includes an opt-in, fail-closed sandbox runner for safer command execution from **Codex, OpenCode, Claude Code, and Pi**. It gives all four agents one portable capability contract with four least-privilege profiles, while using native Seatbelt isolation on macOS and Bubblewrap on Linux.
  45. - **Lightweight:** one contained command tree, with no container image, daemon, or persistent sandbox to manage.
  46. - **Portable:** the same CLI and skill policy work from all four supported coding agents.
  47. - **Least privilege:** skills default to `observe`, `develop`, `network-read`, or `publish`, with explicit escalation checks.
  48. - **Fail closed:** unavailable isolation, undeclared escalation, missing confirmation, or backend failure stops execution instead of retrying outside the sandbox.
  49. Install the CLI, verify the boundary from the active agent session, and run a command with the skill's trusted profile:
  50. ```bash
  51. npm install --global @waybarrios/opencode-power-pack@0.6.0
  52. opencode-power-pack sandbox doctor --json
  53. opencode-power-pack sandbox exec --skill code-review -- git status --short
  54. ```
  55. The current guarantee is intentionally precise: the command and its descendants are `shell-contained`. Automatic host routing and whole-agent isolation are not claimed yet. See [run the sandbox from each agent](#run-the-sandbox-from-each-agent) and the [complete compatibility specification](docs/sandbox-compatibility.md).
  56. ## Claude Code Quick Install
  57. Run these commands inside Claude Code:
  58. ```text
  59. /plugin marketplace add waybarrios/opencode-power-pack
  60. /plugin install opencode-power-pack@opencode-power-pack
  61. ```
  62. Run `/reload-plugins` if the installation summary requests it. Skills are namespaced to avoid collisions, so explicit invocations use names such as `/opencode-power-pack:code-review` and `/opencode-power-pack:feature-dev`.
  63. ## Codex Quick Install
  64. ```bash
  65. codex plugin marketplace add waybarrios/opencode-power-pack --ref main
  66. codex plugin add opencode-power-pack@opencode-power-pack
  67. ```
  68. Start a new Codex session, open `/plugins` to confirm the installation, or invoke a workflow explicitly with `$code-review`, `$feature-dev`, `$security-review`, and the other bundled skills.
  69. ## Pi Quick Install
  70. ```bash
  71. pi install git:github.com/waybarrios/opencode-power-pack
  72. ```
  73. Pi discovers the fifty-five skills declared by the package. Use `pi list` to verify the installation. For a project-local installation recorded in `.pi/settings.json`, add `-l`:
  74. ```bash
  75. pi install git:github.com/waybarrios/opencode-power-pack -l
  76. ```
  77. ## Selective Install With npm
  78. Install only the workflows you want into the shared `.agents/skills` location recognized by Codex, OpenCode, and Pi:
  79. Use `npx` when you want to run the latest installer without keeping a global copy:
  80. ```bash
  81. # Inspect every profile and skill
  82. npx @waybarrios/opencode-power-pack list
  83. # Install the balanced recommended profile for your user
  84. npx @waybarrios/opencode-power-pack install --profile recommended
  85. # Install individual skills
  86. npx @waybarrios/opencode-power-pack install code-review security-review
  87. # Install a profile only for the current repository
  88. npx @waybarrios/opencode-power-pack install --profile review --project
  89. # Preview a selection without writing anything
  90. npx @waybarrios/opencode-power-pack install --profile security --dry-run
  91. # Update previously copied skills from the latest package
  92. npx @waybarrios/opencode-power-pack@latest install --profile review --force
  93. ```
  94. Or install the package globally and use its shorter executable:
  95. ```bash
  96. npm install --global @waybarrios/opencode-power-pack
  97. opencode-power-pack list
  98. opencode-power-pack install --profile recommended
  99. opencode-power-pack install code-review security-review
  100. ```
  101. To update or remove the global installation:
  102. ```bash
  103. npm update --global @waybarrios/opencode-power-pack
  104. npm uninstall --global @waybarrios/opencode-power-pack
  105. ```
  106. The npm package is scoped as `@waybarrios/opencode-power-pack` to give it an unambiguous registry identity. The package still installs the shorter `opencode-power-pack` executable for direct CLI use. No global installation is required when using `npx`.
  107. ### GitHub Packages mirror
  108. Every GitHub release can also be published from its immutable tag to the repository's GitHub Packages page. npmjs remains the recommended public installation source because it works without GitHub authentication.
  109. To install the GitHub Packages mirror, first authenticate with a GitHub personal access token (classic) that has `read:packages`, then select the GitHub registry explicitly:
  110. ```bash
  111. npm login --scope=@waybarrios --auth-type=legacy --registry=https://npm.pkg.github.com
  112. npm install --global @waybarrios/opencode-power-pack --registry=https://npm.pkg.github.com
  113. opencode-power-pack list
  114. ```
  115. Run the installer directly from the mirror after authentication:
  116. ```bash
  117. npx --registry=https://npm.pkg.github.com @waybarrios/opencode-power-pack list
  118. ```
  119. Maintainers publish the mirror with the `Publish GitHub Package` workflow. Release events publish automatically, while a manual run requires an existing release tag such as `v0.6.0`. The workflow verifies that the tag matches `package.json`, uploads the artifact already verified by required release CI without rerunning lifecycle scripts, and authenticates with the repository-scoped `GITHUB_TOKEN`; no package token is stored in the repository.
  120. | Profile | Intended use |
  121. |---|---|
  122. | `recommended` | Balanced software development without the specialized security and ML catalogs |
  123. | `review` | Comprehensive and focused code review plus quality and differential review |
  124. | `feature-dev` | End-to-end feature workflow with its explorer, architect, and reviewer dependencies |
  125. | `frontend` | Frontend implementation and anti-generic design critique |
  126. | `security` | Security review, threat modeling, static analysis, validation, and reporting |
  127. | `huggingface` | Model discovery, local inference, training, evaluation, Spaces, and AWS deployment |
  128. | `authoring` | Skills, MCP servers, technical-paper summaries, and natural copy editing with `humanizer` |
  129. | `project-memory` | Audit and maintain durable project guidance |
  130. Profiles automatically include required companion skills and can be combined with individual skill names. The default destination is `~/.agents/skills`; `--project` finds the current Git root and installs into its `.agents/skills` directory. Existing directories are skipped unless `--force` is provided, and replacements are staged with rollback plus per-skill locking. Every copied skill retains its provenance, third-party notices, and license texts. Use `--dry-run` to preview or `--all` to install the complete catalog.
  131. To install `humanizer` on its own, or the four-skill `authoring` profile:
  132. ```bash
  133. npx @waybarrios/opencode-power-pack install humanizer --dry-run
  134. npx @waybarrios/opencode-power-pack install --profile authoring
  135. ```
  136. The npm installer and the full Claude Code/Codex/OpenCode plugins are alternative activation paths. Claude Code users should prefer the namespaced plugin because the selective installer currently targets the portable `.agents/skills` location used by Codex, OpenCode, and Pi. If a full plugin is already active, selectively copying the same skills does not reduce that plugin's loaded catalog.
  137. ## Sandbox Capability Contract
  138. The package assigns every bundled skill a default least-capability profile and an explicit set of allowed escalations. Native plugin and package installations expose the skills but do not install the sandbox executable, so install the npm CLI in the same operating-system environment as the agent:
  139. ```bash
  140. npm install --global @waybarrios/opencode-power-pack@0.6.0
  141. opencode-power-pack sandbox doctor
  142. opencode-power-pack sandbox resolve --skill code-review
  143. opencode-power-pack sandbox resolve --sandbox-profile publish --json
  144. opencode-power-pack sandbox exec --skill code-review -- rg TODO .
  145. ```
  146. | Sandbox profile | Workspace | Network and credentials | External side effects |
  147. |---|---|---|---|
  148. | `observe` | Read | Denied | Denied |
  149. | `develop` | Write | Denied | Denied |
  150. | `network-read` | Write | Explicitly approved | Denied |
  151. | `publish` | Write | Explicitly approved | Explicit confirmation required |
  152. Run an allowed network escalation by naming the destination explicitly:
  153. ```bash
  154. opencode-power-pack sandbox exec \
  155. --skill code-review \
  156. --sandbox-profile network-read \
  157. --allow-domain api.github.com \
  158. -- curl https://api.github.com/repos/owner/repository
  159. ```
  160. The literal `--` is required. Everything after it is the contained child command and its arguments. The runner preserves the child exit code and never falls back to an unsandboxed execution.
  161. ### Run the sandbox from each agent
  162. Run `sandbox doctor` inside the same agent session that will execute the command. Success in a separate terminal does not prove the active agent can start the native backend.
  163. | Host | How to invoke the runner |
  164. |---|---|
  165. | Codex | Ask Codex to use its shell tool for `opencode-power-pack sandbox doctor`, then for the complete `sandbox exec` command. |
  166. | OpenCode | Ask OpenCode to use its shell tool for `opencode-power-pack sandbox doctor`, then for the complete `sandbox exec` command. |
  167. | Claude Code | Ask Claude Code to use Bash for `opencode-power-pack sandbox doctor`, then for the complete `sandbox exec` command. |
  168. | Pi | Ask Pi to use its shell tool for `opencode-power-pack sandbox doctor`, then for the complete `sandbox exec` command. |
  169. Use this prompt in Codex, OpenCode, or Pi:
  170. ```text
  171. Use the shell tool to run `opencode-power-pack sandbox doctor`.
  172. If it succeeds, run `opencode-power-pack sandbox exec --skill code-review -- git status --short`.
  173. Do not run the child command separately and do not retry without the sandbox.
  174. ```
  175. For Claude Code, use the same prompt with `Use Bash` in the first sentence. Keep the entire `sandbox exec` invocation in one host shell call.
  176. ### Platform requirements
  177. The runner requires Node.js 20.11.0 or newer.
  178. | Platform | Status | Additional requirements |
  179. |---|---|---|
  180. | macOS | Supported | Native Seatbelt; `sandbox doctor` performs an operational preflight. |
  181. | Linux | Supported | Trusted system installations of `bubblewrap`, `socat`, and `rg`, plus working unprivileged user and network namespaces. |
  182. | WSL2 | Expected, dedicated CI pending | The same Linux dependencies and namespace support. |
  183. | WSL1 | Unsupported | Required namespace primitives are unavailable. |
  184. | Windows native | Fails closed | Platform-specific provisioning and boundary tests are not complete. |
  185. The contract continues to report `advisory` because skill metadata does not enforce itself. The npm package also provides an opt-in `sandbox exec` runner backed by a pinned native runtime. When its operational preflight succeeds, `sandbox doctor` reports `shell-contained`; `Strict ready` remains `no` until a host adapter blocks ordinary tool bypasses.
  186. For `network-read`, the runner enforces conventional HTTP read methods (GET, HEAD, and OPTIONS). This is a best-effort side-effect boundary because an approved endpoint can still implement unsafe GET behavior. Use narrow destination grants and expose only credentials required for the retrieval.
  187. The runner is lightweight and on demand: it contains one command tree with native OS primitives, starts no persistent container or daemon, denies unrelated host reads, and deletes its private temporary directory after execution. `sandbox doctor` exits nonzero when the native boundary is unavailable.
  188. See [Sandbox Compatibility Across Coding Agents](docs/sandbox-compatibility.md) for the current Codex, Claude Code, OpenCode, and Pi matrix, platform requirements, activation design, fail-closed behavior, and adapter roadmap.
  189. Selective installations include a generated `SANDBOX_POLICY.json` beside each copied `SKILL.md`. The file is self-describing and preserves the skill's default profile even when the central package is not present. It remains advisory until an enforcement backend and host adapter are active.
  190. ## Why This Exists
  191. Claude Code, Codex, OpenCode, and Pi read `SKILL.md` workflows, but many valuable workflows originated as Claude-specific commands and agents. Copying those artifacts directly does not preserve their orchestration, permissions, or subagent behavior.
  192. This package exposes the shared skills as a namespaced Claude Code plugin, registers feature-development specialist roles as read-only OpenCode subagents, lets Codex execute the same phase assignments with its native subagent workflow, and exposes all fifty-five skills as a Pi package. It ships immutable provenance for every upstream work.
  193. It complements [obra/superpowers](https://github.com/obra/superpowers), which provides process skills such as brainstorming, TDD, debugging, and plan execution.
  194. ## What's Inside
  195. | Category | Skill | Purpose |
  196. |---|---|---|
  197. | Review | `code-review` | Multi-agent PR review with confidence filtering and reproduction scenarios |
  198. | Review | `security-review` | Security-focused review with category coverage and exploit-path validation |
  199. | Review | `supply-chain-risk-auditor` | Flag dependencies at heightened risk of exploitation or takeover |
  200. | Review | `sharp-edges` | Identify error-prone APIs and footgun configurations before they cause security mistakes |
  201. | Review | `insecure-defaults` | Detect fail-open insecure defaults (hardcoded secrets, weak auth, permissive config) |
  202. | Review | `fp-check` | Systematically verify a suspected security bug to a TRUE/FALSE POSITIVE verdict |
  203. | Review | `vuln-report` | Draft a single-vulnerability disclosure report in GitHub advisory style |
  204. | Review | `agentic-actions-auditor` | Audit GitHub Actions workflows for prompt-injection risk in AI agent integrations |
  205. | Review | `security-threat-model` | Model threats using repository evidence, trust boundaries, assets, and abuse paths |
  206. | Review | `differential-review` | Security-focused differential review of a diff/PR with blast-radius analysis |
  207. | Review | `variant-analysis` | Find similar vulnerabilities/bugs across a codebase from one initial pattern |
  208. | Review | `sarif-parsing` | Parse, filter, dedupe, and convert SARIF output from CodeQL/Semgrep/other scanners |
  209. | Review | `semgrep` | Run a Semgrep static analysis scan with approval-gated ruleset selection |
  210. | Review | `semgrep-rule-creator` | Write custom Semgrep rules for a specific vulnerability/bug pattern |
  211. | Review | `semgrep-rule-variant-creator` | Port an existing Semgrep rule to additional target languages |
  212. | Review | `codeql` | Run a CodeQL scan using interprocedural data flow and taint tracking |
  213. | Feature development | `feature-dev` | Seven-phase workflow from discovery through implementation and review |
  214. | Feature development | `code-explorer` | Trace a feature across entry points, layers, state, and dependencies |
  215. | Feature development | `code-architect` | Produce a file-level architecture and implementation blueprint |
  216. | Feature development | `code-reviewer` | Adversarial review of a focused local change set |
  217. | Design | `frontend-design` | Create distinctive interfaces with an accessibility and craft rubric |
  218. | Design | `ai-slop` | Rubric to judge whether a design is generic "AI slop" or genuinely product-fit |
  219. | Code quality | `code-quality` | Linting, complexity, and review checklists across Rust, TypeScript, Python, and shell |
  220. | Code quality | `design-patterns` | Pattern trade-offs (when to use, when to skip) for Rust, TS/React, and Django/Python |
  221. | Research | `paper-summarizer` | Extract actionable findings and a claim-evidence map from academic/technical papers |
  222. | Authoring | `humanizer` | Rewrite AI-generated copy in ES/EN into direct, natural prose without inventing facts |
  223. | Authoring | `mcp-builder` | Design and build MCP servers in TypeScript or Python |
  224. | Authoring | `skill-creator` | Create, test, and improve reusable `SKILL.md` workflows |
  225. | Hugging Face | `hf-cli` | Use the `hf` CLI for auth, cache, repos, jobs, papers, and Spaces |
  226. | Hugging Face | `hf-mem` | Estimate memory needed to load Safetensors/GGUF weights for inference |
  227. | Hugging Face | `huggingface-best` | Find and compare the best model for a task via official benchmark leaderboards |
  228. | Hugging Face | `huggingface-datasets` | Dataset Viewer API workflows for metadata, pagination, search, and filters |
  229. | Hugging Face | `huggingface-papers` | Look up and analyze Hugging Face / arXiv paper pages via the papers API |
  230. | Hugging Face | `huggingface-paper-publisher` | Publish and manage research papers on the Hub, including research-article templates |
  231. | Hugging Face | `huggingface-tool-builder` | Build reusable scripts/tools around the Hugging Face API |
  232. | Hugging Face | `huggingface-local-models` | Select and run GGUF models locally with llama.cpp (CPU/Metal/CUDA/ROCm) |
  233. | Hugging Face training | `huggingface-llm-trainer` | Fine-tune LLMs (TRL: SFT/DPO/GRPO, or Unsloth) on Hugging Face Jobs |
  234. | Hugging Face training | `huggingface-vision-trainer` | Fine-tune object detection, image classification, and SAM/SAM2 segmentation models |
  235. | Hugging Face training | `train-sentence-transformers` | Train bi-encoder/cross-encoder/sparse embedding models |
  236. | Hugging Face training | `trl-training` | Train with the TRL CLI using SFT, DPO, GRPO, KTO, RLOO, and reward modeling |
  237. | Hugging Face training | `huggingface-community-evals` | Run local evaluations with inspect-ai / lighteval |
  238. | Hugging Face training | `huggingface-trackio` | Track and visualize ML training experiments with Trackio |
  239. | Hugging Face Spaces | `huggingface-spaces` | Build, deploy, and debug Gradio/Docker/static Spaces |
  240. | Hugging Face Spaces | `huggingface-gradio` | Build Gradio web UIs and demos in Python |
  241. | Hugging Face Spaces | `huggingface-lora-space-builder` | Build and publish a Gradio Space demo for a user-provided LoRA |
  242. | Hugging Face Spaces | `huggingface-zerogpu` | Write/debug ZeroGPU-constrained Gradio Space code |
  243. | Hugging Face cloud | `hf-cloud-aws-context-discovery` | Discover the user's local AWS context before any AWS work |
  244. | Hugging Face cloud | `hf-cloud-python-env-setup` | Set up an isolated Python env for SageMaker/AWS work |
  245. | Hugging Face cloud | `hf-cloud-sagemaker-deployment-planner` | Entry point for planning a SageMaker model deployment |
  246. | Hugging Face cloud | `hf-cloud-sagemaker-iam-preflight` | Ensure a usable SageMaker execution role exists before deploying |
  247. | Hugging Face cloud | `hf-cloud-serving-image-selection` | Pick the right SageMaker serving container image |
  248. | Hugging Face cloud | `hf-cloud-sagemaker-production-defaults` | Create a SageMaker endpoint with autoscaling, alarms, and tagging by default |
  249. | JavaScript | `transformers-js` | Run Hugging Face models directly in JS/TS (browser or Node/Bun/Deno) |
  250. | Project memory | `agents-md-improver` | Audit project rules and propose targeted improvements |
  251. | Project memory | `agents-md-revise` | Capture durable session learnings in project rules |
  252. `code-explorer`, `code-architect`, and `code-reviewer` are standalone skills and specialist roles used by `feature-dev`. OpenCode registers named least-privilege agents; Codex can carry out the same assignments with native subagents and can use matching custom agents when the user configures them.
  253. ### Which Code Review Skill?
  254. | Situation | Use | Why |
  255. |---|---|---|
  256. | PR, branch, commit range, or complete pending diff | `code-review` (Comprehensive Code Review) | Freezes the scope, runs multiple review lenses, cross-checks candidates, and can post validated findings to GitHub when requested |
  257. | One file, function, or small local change | `code-reviewer` (Focused Code Review) | Uses a lighter two-pass review with full-file and caller context |
  258. | `feature-dev` quality phase | `code-reviewer` (Focused Code Review) | Acts as the focused specialist role dispatched by the feature workflow |
  259. They are complementary rather than duplicate workflows. Codex shows distinct user-facing titles, Claude Code namespaces both under `opencode-power-pack`, and the stable skill IDs continue to work across every host. Start with `code-review` for merge decisions and `code-reviewer` for focused iteration.
  260. ## Installation
  261. ### Prerequisites
  262. - Git
  263. - Node.js 20.11.0 or newer, including npm/npx, for [selective installation](#selective-install-with-npm) and the sandbox runner
  264. - One supported host:
  265. - A current Claude Code installation with plugin support: <https://code.claude.com/docs/en/plugins>
  266. - OpenCode 1.18.7 or newer, including OpenCode 2: <https://opencode.ai>
  267. - A current Codex CLI or Codex desktop environment with plugin support: <https://developers.openai.com/codex/>
  268. - A current Pi coding agent installation: <https://pi.dev/>
  269. ### Claude Code
  270. Add the repository marketplace and install its plugin from inside Claude Code:
  271. ```text
  272. /plugin marketplace add waybarrios/opencode-power-pack
  273. /plugin install opencode-power-pack@opencode-power-pack
  274. ```
  275. Run `/reload-plugins` if prompted. Claude Code loads plugin skills from the repository-level `skills/` directory and prefixes them with the plugin namespace, preventing collisions with built-in or separately installed skills. For local development, run `claude --plugin-dir .` from a clone of this repository.
  276. ### Codex CLI And Desktop
  277. Add this repository as a marketplace, then install the plugin:
  278. ```bash
  279. codex plugin marketplace add waybarrios/opencode-power-pack --ref main
  280. codex plugin add opencode-power-pack@opencode-power-pack
  281. codex plugin list --marketplace opencode-power-pack
  282. ```
  283. Start a new Codex session after installation so the fifty-five bundled skills are loaded. Use `/plugins` to inspect the installed plugin or `$` to select one of its skills explicitly. Codex plugin packaging follows the [official plugin structure](https://developers.openai.com/plugins/build/plugins).
  284. For a smaller personal or repository-specific set, use the [selective npm installer](#selective-install-with-npm) instead of the full plugin.
  285. ### OpenCode From GitHub
  286. On OpenCode 2:
  287. ```bash
  288. opencode plugin add "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  289. ```
  290. On OpenCode 1:
  291. ```bash
  292. opencode plugin --global "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  293. ```
  294. Restart OpenCode after installation. One plugin adapts to the host version: OpenCode 2 registers all fifty-five skills and the three specialist agents through the V2 plugin API, and OpenCode 1 keeps the original `config.skills.paths` and `config.agent` hook. Either way the skills are invocable by name and the three roles are registered as read-only subagents.
  295. The compatibility smoke tests cover OpenCode **1.18.7, 2.0.14, and 2.0.32**. The `SKILL.md` format stays the same across these hosts.
  296. To pin a release, append its tag to the same spec. OpenCode 2 needs `v0.6.0` or later, because `v0.5.0` predates the V2 port:
  297. ```bash
  298. opencode plugin add "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git#<tag>"
  299. ```
  300. ### OpenCode From A Local Clone
  301. ```bash
  302. git clone https://github.com/waybarrios/opencode-power-pack.git ~/code/opencode-power-pack
  303. opencode plugin add "opencode-power-pack@git+file:///home/you/code/opencode-power-pack"
  304. ```
  305. Use an absolute `file://` URL adjusted for your operating system. The target directory must be a Git repository. On OpenCode 1, replace `plugin add` with `plugin --global` as shown above.
  306. ### Pi From GitHub
  307. ```bash
  308. pi install git:github.com/waybarrios/opencode-power-pack
  309. ```
  310. Pi installs the Git package and loads the `skills/` directory declared in `package.json`. Use `-l` to save it to project settings instead of user settings. See the [Pi package documentation](https://pi.dev/docs/latest/packages) for source, update, filtering, and trust behavior.
  311. Pi can also narrow an installed package through `pi config`. The npm installer is useful when you want the same selected set to be discovered by Pi, Codex, and OpenCode from `.agents/skills`.
  312. ### Verify Claude Code
  313. From a local clone, validate both the plugin and marketplace manifests:
  314. ```bash
  315. claude plugin validate . --strict
  316. ```
  317. After marketplace installation, run `/plugin` and confirm that `opencode-power-pack` is enabled. Invoke `/opencode-power-pack:code-review` for a first test.
  318. ### Verify OpenCode
  319. Restart OpenCode, then check the registered skills and specialist agents.
  320. On OpenCode 2:
  321. ```bash
  322. opencode debug agents
  323. ```
  324. The output should include `code-explorer`, `code-architect`, and `code-reviewer` as `subagent` entries. OpenCode 2 enforces their read-only restrictions when evaluating permission requests, so the agent listing can show the host's base permissions. In the TUI, the slash catalog should list `/code-review`, `/feature-dev`, `/frontend-design`, and the other bundled skills.
  325. On OpenCode 1:
  326. ```bash
  327. opencode debug skill
  328. opencode debug agent code-explorer
  329. ```
  330. The first command should include all fifty-five unprefixed skill names. The second should report a `subagent` with editing denied. In the TUI, `ctrl+p` should list `/code-review`, `/feature-dev`, `/frontend-design`, and the other skill-derived commands.
  331. ### Verify Codex
  332. ```bash
  333. codex plugin list --marketplace opencode-power-pack
  334. ```
  335. The list should show `opencode-power-pack` as installed. Start a new session and explicitly select `$code-review`, `$feature-dev`, or another bundled skill for a first test.
  336. ### Verify Pi
  337. ```bash
  338. pi list
  339. ```
  340. The package list should include the Git source for `waybarrios/opencode-power-pack`. Start a new Pi session after installation so its bundled skills are available.
  341. ## Updating
  342. For Claude Code:
  343. ```text
  344. /plugin marketplace update opencode-power-pack
  345. /plugin update opencode-power-pack@opencode-power-pack
  346. ```
  347. Run `/reload-plugins` if prompted.
  348. For Codex:
  349. ```bash
  350. codex plugin marketplace upgrade opencode-power-pack
  351. codex plugin add opencode-power-pack@opencode-power-pack
  352. ```
  353. Start a new session after the upgrade.
  354. For Pi:
  355. ```bash
  356. pi update git:github.com/waybarrios/opencode-power-pack
  357. ```
  358. For an OpenCode GitHub installation on OpenCode 2:
  359. ```bash
  360. opencode plugin update "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  361. ```
  362. On OpenCode 1, reinstall with the same spec:
  363. ```bash
  364. opencode plugin --global "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  365. ```
  366. Restart OpenCode after the command finishes. Pass the exact configured package specification, including any pinned tag. Local-clone users should run `git pull` in the clone, then update using their original `opencode-power-pack@git+file://...` specification on OpenCode 2, or reinstall it on OpenCode 1.
  367. ## Uninstalling
  368. For Claude Code:
  369. ```text
  370. /plugin uninstall opencode-power-pack@opencode-power-pack
  371. ```
  372. For Codex:
  373. ```bash
  374. codex plugin remove opencode-power-pack@opencode-power-pack
  375. ```
  376. For Pi:
  377. ```bash
  378. pi remove git:github.com/waybarrios/opencode-power-pack
  379. ```
  380. For OpenCode 2, remove the plugin from the global configuration:
  381. ```bash
  382. opencode plugin remove "opencode-power-pack@git+https://github.com/waybarrios/opencode-power-pack.git"
  383. ```
  384. Use your exact configured package specification if it differs from this example. Alternatively, delete the `opencode-power-pack@...` entry from the `plugins` array in the config where it was installed, then restart OpenCode. On OpenCode 1, delete the entry from the `plugin` array instead. There are no command symlinks or copied command files to remove.
  385. ## Troubleshooting
  386. | Symptom | Cause | Fix |
  387. |---|---|---|
  388. | Claude Code cannot add the marketplace | The checkout predates `.claude-plugin/marketplace.json` or the repository is inaccessible | Update the checkout, verify repository access, and run `claude plugin validate . --strict` locally |
  389. | Claude Code installed the plugin but a skill name conflicts | The unprefixed skill came from another source | Invoke the namespaced form, such as `/opencode-power-pack:code-review` |
  390. | Codex cannot find the marketplace | Marketplace snapshot is missing or stale | Run `codex plugin marketplace add waybarrios/opencode-power-pack --ref main`, or upgrade the existing marketplace |
  391. | Codex installed the plugin but skills do not appear | The current session predates installation | Start a new Codex session and inspect `/plugins` |
  392. | Pi does not show the skills | Package is absent, disabled, or the session predates installation | Run `pi list`, inspect `pi config`, and start a new Pi session |
  393. | Skills or commands do not appear | OpenCode is older than 1.18.7 | Upgrade OpenCode, restart, and check the slash catalog or run `opencode debug skill` |
  394. | Skills are missing from the catalog | Plugin installation failed or its `plugins` entry is absent | Reinstall with `opencode plugin add <module-spec>` (OpenCode 2) or `opencode plugin --global <module-spec>` (OpenCode 1) and inspect the reported error |
  395. | Specialist agents are missing | Stale plugin checkout | Update the plugin, restart, and run `opencode debug agents` (OpenCode 2) or `opencode debug agent code-explorer` (OpenCode 1) |
  396. | Installation reports a Git error | Invalid URL, network failure, or inaccessible repository | Validate the source with `git ls-remote <url>` |
  397. | A workflow is rushed or incomplete | The backing model skipped multi-stage instructions | Use a stronger model and inspect whether required subagent tools are available |
  398. ## Invocation
  399. Claude Code exposes plugin skills as `/opencode-power-pack:<skill>`, Codex uses `$` mentions, OpenCode exposes the bundled skills by name (`/` catalog on OpenCode 2, same-named slash commands on OpenCode 1), and Pi loads workflows from the installed skills package:
  400. ```text
  401. /code-review
  402. /security-review
  403. /feature-dev
  404. /code-explorer
  405. /code-architect
  406. /code-reviewer
  407. /frontend-design
  408. /mcp-builder
  409. /skill-creator
  410. /agents-md-improver
  411. /agents-md-revise
  412. ```
  413. Examples:
  414. ```text
  415. $code-review review the current branch against main
  416. $feature-dev add a logout button to the topbar
  417. /code-review --comment
  418. /code-review review https://github.com/owner/repo/pull/449
  419. /feature-dev add a logout button to the topbar
  420. /security-review
  421. /frontend-design pricing page, brutalist tone, single-screen
  422. /opencode-power-pack:code-review
  423. /opencode-power-pack:feature-dev add a logout button to the topbar
  424. ```
  425. In OpenCode, an explicit command with the same name takes precedence over a skill-derived command.
  426. ## How It Works
  427. ```text
  428. opencode-power-pack
  429. |
  430. +-- .claude-plugin/
  431. | +-- plugin.json declares the namespaced Claude Code plugin
  432. | +-- marketplace.json makes the repository installable in Claude Code
  433. |
  434. +-- .codex-plugin/plugin.json
  435. | +-- packages all skills/ for Codex and compatible plugin surfaces
  436. |
  437. +-- .agents/plugins/marketplace.json
  438. | +-- exposes the repository as an installable Codex marketplace
  439. |
  440. +-- package.json
  441. | +-- declares skills/ as a Pi package for pi install
  442. |
  443. +-- sandbox/contract.json
  444. | +-- maps all fifty-five skills to versioned capability profiles
  445. | +-- records allowed escalation paths without claiming enforcement
  446. |
  447. +-- bin/sandbox/policy.mjs
  448. | +-- validates, resolves, and reports the shared capability contract
  449. |
  450. +-- bin/sandbox/runtime.mjs
  451. | +-- runs opt-in commands in the pinned native sandbox backend
  452. |
  453. +-- .opencode/plugins/opencode-power-pack.js
  454. | +-- OpenCode 1: registers config.skills.paths and config.agent
  455. | +-- OpenCode 2: registers skills and agents through the V2 transforms
  456. | +-- registers code-explorer as a read-only subagent
  457. | +-- registers code-architect as a read-only subagent
  458. | +-- registers code-reviewer with read-only Git commands
  459. |
  460. +-- .opencode/lib/agent-config.js
  461. | +-- derives skill and subagent definitions from the SKILL.md frontmatter
  462. | +-- builds the OpenCode 1 permission maps and OpenCode 2 rulesets
  463. |
  464. +-- skills/<name>/SKILL.md
  465. | +-- shared Codex/OpenCode workflow
  466. | +-- host-native invocation metadata
  467. | +-- immutable source and license metadata
  468. |
  469. +-- UPSTREAMS.json
  470. +-- repository, commit, path, blob, date, and adaptation type
  471. ```
  472. Each `SKILL.md` is the single source for its workflow. Claude Code loads the directory through its plugin namespace. The OpenCode plugin adapts to the host version and derives specialist-agent prompts from those files at startup. Codex loads the same skills from its plugin and follows `feature-dev`'s specialist assignments with its native subagent workflow; installing a skill does not itself create a named custom agent.
  473. The packaged agents deny edits, external network access, and nested subagents. `code-reviewer` additionally allows a narrow set of read-only Git commands.
  474. On OpenCode 2, permission evaluation enforces these restrictions even when global rules allow broader access. User-configured agents with their own system prompt or a primary mode replace the bundled role and keep the host's permissions. Explicit user denials and approval prompts still apply to bundled agents.
  475. ## Scope And Non-Goals
  476. | In scope | Out of scope |
  477. |---|---|
  478. | Claude Code, Codex, OpenCode, and Pi skills and invocation | Whole-agent, browser, connector, or desktop isolation |
  479. | Portable and host-native subagent orchestration | Proprietary or non-redistributable plugins |
  480. | Licensed adaptations with immutable provenance | Automatic trust of third-party skill catalogs |
  481. | Explicit permission boundaries and regression tests | Supporting OpenCode versions older than 1.18.7 or obsolete Codex plugin formats |
  482. | Versioned sandbox contracts and an opt-in shell-contained runner | Claiming whole-agent containment before host adapters block bypass tools |
  483. ## Contributing
  484. Contributions are welcome for:
  485. - Improvements based on reproducible workflow failures
  486. - Behavioral evaluations and adversarial fixtures
  487. - Portable references and deterministic helper scripts
  488. - New ports with a verified license and immutable upstream source
  489. Before opening a PR:
  490. ```bash
  491. npm test
  492. npm run test:sandbox
  493. npm run smoke:opencode
  494. npm pack --dry-run
  495. ```
  496. Live behavioral evaluation is opt-in, needs `OPENCODE_EVAL_MODEL`, and runs only from a source checkout:
  497. ```bash
  498. OPENCODE_EVAL_MODEL=provider/model npm run eval:behavioral
  499. ```
  500. See [`evals/behavioral/README.md`](evals/behavioral/README.md) for the replay, review, and acceptance workflow. Snapshots are content-addressed evidence from one reviewed model execution, not universal guarantees.
  501. Every skill must use a lowercase hyphenated directory/name, provide a trigger-specific description, remain under 500 lines where practical, and record its exact upstream in `UPSTREAMS.json`.
  502. ## Acknowledgments
  503. The bundled skills are modified upstream works. This repository contributes their Claude Code, Codex, OpenCode, and Pi packaging, additional workflow guidance, tests, and compatibility layer.
  504. | Upstream | Pinned source | Used for |
  505. |---|---|---|
  506. | Anthropic | [`claude-plugins-official@bdca23e8`](https://github.com/anthropics/claude-plugins-official/tree/bdca23e8e46f8832d0030c05804ae207786ae37f) | Code review, feature development, frontend design, and project memory |
  507. | Anthropic | [`skills@5128e186`](https://github.com/anthropics/skills/tree/5128e1865d670f5d6c9cef000e6dfc4e951fb5b9) | MCP Builder and Skill Creator |
  508. | Anthropic | [`claude-code-security-review@0c6a49f1`](https://github.com/anthropics/claude-code-security-review/tree/0c6a49f1fa56a1d472575da86a94dbc1edb78eda) | Security Review |
  509. | Jesse Vincent | [`superpowers@6efe32c9`](https://github.com/obra/superpowers/tree/6efe32c9e2dd002d0c394e861e0529675d1ab32e) | OpenCode plugin registration pattern |
  510. See [`UPSTREAMS.json`](UPSTREAMS.json) for exact source paths and Git blobs. See [`THIRD_PARTY_NOTICES.md`](THIRD_PARTY_NOTICES.md) for attribution and modification notices.
  511. ## License
  512. The wrapper code and original project material are MIT-licensed under [LICENSE](LICENSE). Modified skills derived from Anthropic's official plugins and skills are Apache-2.0, except `security-review`, whose upstream is MIT. See [THIRD_PARTY_NOTICES.md](THIRD_PARTY_NOTICES.md), [UPSTREAMS.json](UPSTREAMS.json), and [LICENSES/](LICENSES/) for exact terms.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.