xdg-claude

CLI tool skills that minimize context usage through targeted extraction instead of reading entire files

5 skills · hooks

Surgical code refactoring agents for analyzing and improving code quality without changing behavior

4 skills · 2 agents

Code review agent for analyzing code quality, security, performance, and maintainability

2 skills · 1 agent

Audit code comments for information content and clarity against Philosophy of Software Design and The Art of Readable Code, returning exact patches tagged mechanical or judgment for the caller to apply

2 skills · 1 agent

Git commit agent with intelligent change analysis and commit message conventions

2 skills · 1 agent

Execute focused work in clean, isolated contexts without polluting the main conversation history. Ideal for exploratory analysis, parallel work streams, and context-heavy operations.

2 skills · 1 agent

Implement TODO.md items via isolated subagents with adversarial verification, code review, human-in-the-loop checks, and per-subsection commits

1 skill

Consult OpenAI Codex headlessly for an independent review of a doc, plan, diff, or code, then reconcile its findings with Claude's own view

1 skill

Consult any opencode-configured model headlessly for an independent review of a doc, plan, diff, or code, resolving short model names like 'astra', then reconcile its findings with Claude's own view

1 skill

Injects writing style guidance at session start to avoid common AI writing patterns

hooks

Injects coding best practices from Philosophy of Software Design and The Art of Readable Code at session start and on subagent dispatch

hooks

Eval fixture: one skill with a soft trigger, used to measure whether prioritize-skills raises skill invocation.

1 skill

Record durable engineering decisions in an append-only decision log, with a fixed entry template and consent before creating the log in an unfamiliar repo

1 skill

Draft, amend, and close out the document ladder for software work: PRD, milestone system design, inchstone behavioral spec, and the living docs they feed

1 skill

Injects writing style guidance from The Elements of Style at session start

hooks

Enforce golangci-lint: run on Stop, block git commit on failure, block //nolint directives and edits to .golangci.yml

hooks

Conducts a structured, persisted, tree-shaped interview to externalize an under-formed plan, design, or strategy

1 skill

Manage Jira issues from the command line via jira-cli.

1 skill

Injects pithy, detailed communication style guidance at session start for high signal-to-noise ratio

hooks

Reference skill for authoring Claude Code plugins: component types, the three skill patterns, frontmatter, agents, hooks, manifests, and worked examples

1 skill

Injects a directive at session and subagent start that requires checking for and invoking applicable skills before responding or acting

hooks

Guided, interactive walkthrough of a file or diff one unit at a time, driven by a persisted checklist so the review spans sessions

1 skill

Create and maintain structured TODO.md implementation plans with phased tasks, a testing philosophy, and verification checklists

1 skill