By DevGuyRash
Audit a skill's packaging, trigger behavior, task leverage, context design, verification loop, and AGENTS.md adherence, then produce a concise Skill Improvement Brief. Use when the task involves: (1) Auditing or self-auditing a skill for quality or correctness, (2) Health-checking a skill's trigger reliability or context efficiency, (3) Evaluating whether a skill belongs as a skill, AGENTS.md entry, explicit prompt, or tool, (4) Reviewing or revising a skill's packaging, descriptions, or references, or (5) Any task requiring structured skill evaluation with concrete improvement checks.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
This repository contains portable agent tooling: dual-host plugin packages, host-agnostic skill payloads, Rust-backed launchers, and repo harness scripts.
Top-level plugins/ contains plugin packages that may bundle skills, hooks,
MCP servers, apps, and host manifests.
Top-level skills/ is kept with .gitkeep for future standalone,
host-agnostic skill packages that are not distributed as plugins.
WHEN adding a plugin package to this repository THEN you SHALL place it under
plugins/<plugin-name>/.
WHEN adding reusable skill content that is not a plugin package THEN you SHALL
place it under skills/<skill-name>/.
WHEN a plugin bundles skill instructions THEN you SHALL keep those bundled
skills inside the plugin package's own skills/ directory.
Marketplace manifests:
.agents/plugins/marketplace.json.claude-plugin/marketplace.jsonCurrent local plugins:
plugins/code-review/plugins/docker-architect/plugins/espanso-dynamic-forms/plugins/excel-foundry/plugins/friction-diagnostics/plugins/gitops-workflow/plugins/goal-foundry/ exposes goal-foundry for both Codex and Claude and
bundles the agnostic $authoring-goals skill payload.plugins/playwright-testing/plugins/project-harness/plugins/rust-development/plugins/skill-auditor/code-reviewUnified code review skill with two workflows:
Both workflows coordinate artifacts under .local/reports/code_reviews/{YYYY-MM-DD}/ and use a bundled mpcr tool for deterministic reviewer/session operations (ID generation, locking, session JSON updates, report writing).
Path: plugins/code-review/skills/code-review/
Migration note: perform-code-review and apply-code-review were consolidated into code-review. Update any tooling or docs that reference skills/perform-code-review/ or skills/apply-code-review/ to use plugins/code-review/skills/code-review/.
docker-architectDeterministic Docker architecture skill spanning both Compose/Swarm deployment design and image supply-chain planning with strict output ordering and traceability IDs (AC-*, IMG-*, RSK-*, O-*).
plugins/docker-architect/skills/docker-architect/scripts/docker-architect-compose (packaged-binary launcher)plugins/docker-architect/skills/docker-architect/scripts/docker-architect-image (packaged-binary launcher)Path: plugins/docker-architect/skills/docker-architect/
scripts/plugin_port.py converts Codex and Claude Code plugin packages and
marketplaces while preserving source trees and writing a conversion report to
.plugin-portability/report.json.
Common commands:
python3 scripts/plugin_port.py inspect <path> --format json|md [--from codex|claude]python3 scripts/plugin_port.py convert <plugin-dir> --to codex|claude --out <output-dir> --mode strict|best-effort [--summary full|json|md]python3 scripts/plugin_port.py convert-marketplace <marketplace-root-or-json> --to codex|claude --out <output-dir> [--summary full|json|md]python3 scripts/plugin_port.py validate <plugin-dir> --host codex|claude [--require-external-validator] [--summary full|json|md]python3 scripts/plugin_port.py roundtrip <plugin-dir> --to codex|claude --tmp <work-dir> [--summary full|json|md]Compatibility contract:
CLAUDE.md files when
targeting Claude; Claude LSP/output styles/themes/monitors/bin/settings when
targeting Codex. Root CLAUDE.md files are moved to
.plugin-portability/preserved/CLAUDE.md in Claude output because Claude
plugin validation rejects plugin-root context files.unsupported, preserved_only, and executable_surfaces.
Invalid skill, command, or agent frontmatter is repaired with generated target
metadata only in best-effort conversion; validation still rejects malformed
source frontmatter.Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub devguyrash/agent-tooling --plugin skill-auditorRun recursive, machine-first code review, review application, and full-cycle convergence workflows with the `mpcr` CLI. Use when the task involves: (1) Reviewing a diff, pull request, or set of commits, (2) Applying structured review findings back to code, (3) Validating canonical review artifacts or report trees, (4) Driving review→apply→verify convergence loops, or (5) Any task requiring systematic, multi-pass code review with per-agent report trees.
Generate hardened, production-ready Docker architecture including Dockerfiles, Compose stacks, and Swarm deploy configs. Use when the task involves: (1) Writing or improving a Dockerfile or multi-stage build, (2) Containerizing an application (Python, Node.js, Rust, Go, Nginx, or custom stacks), (3) Creating or modifying compose.yaml or Docker Swarm deployments, (4) Hardening container security, healthchecks, secrets management, or resource limits, (5) Setting up CI/CD container scanning or image supply-chain controls (hadolint, trivy, docker scout, SBOM, provenance), or (6) Any task involving Docker, containers, or container orchestration.
Build portable dynamic Espanso forms where a script or binary generates form layouts at runtime. Use when the task involves: (1) Creating or modifying Espanso form triggers or match files, (2) Building dynamic Espanso forms with runtime-generated layouts or choices, (3) Scaffolding a new Espanso form provider pattern, (4) Configuring Espanso YAML for scripted fields or runtime-generated options, (5) Debugging or linting Espanso form configurations, (6) Implementing clipboard or output patterns for Espanso expansions, or (7) Any task involving Espanso text expansion with dynamic or scripted form UIs.
Inspect, create, diff, document, audit, automate, and sync Excel workbook artifacts across package-readable workbooks, safe copies, installed desktop Excel, Microsoft Graph workbook APIs, Office Scripts/Add-in hosts, and Fabric/Power BI semantic models. Use when the task involves: (1) workbook lifecycle or direct commands such as inspect/diagnose/capabilities/create/diff/save-as/ convert/repair/compatibility/document-inspect, sheet/table/range/cell/name edits, links, safe export, formulas, calc, Solver, scenarios, forecasts, data tables, or automation artifacts, (2) package inventory or preservation for charts, pivots, slicers, timelines, Power Query, connections, XML/custom XML, OLE, signatures, encryption, sensitivity, VBA/package metadata, or embedded media, (3) desktop Excel host operations, (4) Graph, Office Scripts, Office.js/Add-in, Fabric, Power BI, DAX, TMDL, PBIP, or semantic-model operations, or (5) manifest/artifact synchronization, audit bundles, parity reports, and portable workbook manifests.
REQUIRED for every task. Log every error, failure, unexpected outcome, or friction encountered — no exceptions. Covers: (1) Any tool, command, script, or build step that fails or produces unexpected results, (2) Any instruction, workflow, or environment that behaves differently than expected, (3) Code bugs, compilation errors, test failures, or runtime exceptions, (4) MCP server, subagent, or integration misbehavior, (5) User-requested friction or diagnostic reporting, (6) Any problem, error, or divergence from expected behavior during task execution. If anything goes wrong, it gets logged here. Do not skip this skill.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.