Developer Tools plugins

Find, inspect, and implement UI components from Col's curated catalogue inside your project, with source-linked setup and explicit gaps.

1 skill · 1 MCP server

Ontology-first knowledge packs: claims with verifiable sources (frozen captures checked by sha256 and quote) and a ledger-backed stopping gate.

1 skill · 1 command

Connects the Meridian MCP server and teaches Claude to use the Meridian CLI, add the SDK to a Shopify app and deploy it on Meridian hosting.

1 skill · 1 MCP server

Resolves software trade-offs with TRIZ: the contradiction matrix and 40 inventive principles, the separation principles, and ARIZ-85C walked part by part.

2 skills

Vets sources, collapses copies to independent origins, and grades each claim by how many independent origins support it.

2 skills

SaaS billing best practices. Patterns for subscription lifecycle, failed payments, proration, multi-currency, tax compliance, plan changes, invoicing, and launch checklists.

1 skill

Smart git commits with conventional commit message generation

1 command

Create, test, and bulletproof Claude Code skills using TDD methodology. Based on obra/superpowers writing-skills

1 skill

Persistent Julia REPL sessions for AI agents, sandboxed with bubblewrap

1 MCP server

Run and schedule AI coding agents in isolated git worktrees with the Agent Factory (af) CLI

1 skill

Build a new, high-quality Claude Agent Skill from scratch — frontmatter, tight SKILL.md, progressive-disclosure reference files, a self-test, and packaging as a .skill or plugin. Also audits and improves existing skills and fixes descriptions that don't trigger.

1 skill

Maps a project's dependency graph (JS/TS, Python, Rust, Dart): an Obsidian vault for humans and a compact PROJECT_MAP.md for token-efficient AI navigation, with an auto-refresh hook.

1 skill

Cross-platform secrets workflow for macOS, Windows, and Linux. The `keys` CLI routes values to explicit local sinks without returning plaintext in normal agent tool output. It also organizes local folders and explicit project scopes so approved workers receive only assigned entries and contributors can submit new entries without editing existing ones. This reduces accidental transcript exposure; it is not an isolation boundary against arbitrary code running as the same OS user.

1 skill

RelayPact: bounded, evidence-driven delegation between AI agents.

1 skill

Codex companion for Previa, the AI-first IDE for QA and API testing workflows.

1 skill

Plan reviews of local changes or PRs, then run an approved plan with specialized review agents and evidence-based reporting.

1 skill

Checks for Codex Desktop MSIX updates and keeps the local updater plugin current from GitHub.

1 skill

Community package for Chrome DevTools MCP. Inspect pages, automate flows, analyze performance, and debug browser state.

1 MCP server

Train, evaluate, and extend robots locally through Codex and Apple Silicon.

1 skill

Guarded session-local pointer control and screenshots for Hyprland on Wayland.

1 skill

Reusable backend engineering skill for implementation, debugging, refactoring, and code review with progressive domain references.

1 skill

HAPI collaboration between blue development and green protected-service AI agents.

1 skill

Project-bound agent roles, context, memory and verified workflows through the installed Crewloom CLI.

1 skill

Initialize or migrate an Agent project contract (AGENTS.md, Project Profile, root launcher); ships with project-orchestrator, agent-pack and opinion-manager.

1 skill

Manage Agent behavior rules through templates, partial selection, blank iteration, and authorized-memory candidates, with complete previews and exact version locks.

1 skill

Compares another project against yours and returns a ranked, evidence-backed list of what it has that you lack. Builds nothing until you pick items.

1 skill

Use the Iris CLI to create a local messaging identity, exchange invites, and send or receive encrypted messages with humans and agents. No provider signup, email, or phone number is required.

1 skill

Separates the cognitive load a task needs from the load a code base's structure adds, and names the change that removes the second.

1 skill

Selbst gehostete Remote-Development-Workbench im Browser: Editor, PTY-Terminals, KI-CLIs, Orbit-Canvas und Tech-News, privat via Tailscale.

Jev-powered gates, injection screen and done-check