Agent Skills by Tarragon
Showing 49 vetted skills indexed across 4 GitHub repositories.
tarrragoncompositional-writing
Creates bite-sized, indexable text units for reusable technical documentation and communication.
tarrragonfrontend-with-playwright
Validate frontend DOM topology, layout, and accessibility with Playwright.
tarrragonwrap-decision
Guide teams through structured decision-making with evidence, options, and guardrails.
tarrragonrequirement-protocol
Locate SKILL.md files and extract frontmatter metadata into a YAML block.
tarrragonFrontend with Playwright — SKILL 入口
Validate frontend DOM/CSS/JS interactions using Playwright across vanilla and frameworks.
tarrragonworktree
Automates Git worktree management for ticket-based development with branch safety checks and optional dry-run support.
tarrragonspec
Convert ambiguous tickets into testable feature specifications with /spec init and /spec validate.
tarrragonzellij
Coordinate actions across multiple zellij panes from Claude Code.
tarrragonbroken-link-check
Scan Markdown files in .claude and report broken path references.
tarrragondoc
Organize and query proposals, specs, and use cases via YAML-frontmatter CLI commands.
tarrragontdd
Guide teams through TDD phases with planning, design, and execution logs.
tarrragonagent-team
Coordinate multi-agent Claude teams with ticket-driven workflows and lifecycle stages.
tarrragoni18n-checker
Scan codebases for hard-coded Chinese strings and generate ARB key suggestions.
tarrragonbulk-evaluate
Split large tasks into sub-tickets for independent agent analysis.
tarrragondecision-tree-helper
Transform task descriptions into decision-tree dispatch workflows with complexity scores and agent recommendations.
tarrragonskill-design-guide
Document and standardize Claude Skill design with YAML frontmatter rules.
tarrragonmermaid-ascii
Convert Mermaid diagrams to ASCII or Unicode renderings via CLI.
tarrragondoc-flow
Coordinate project documentation workflows across changelog, worklog, and ticket systems.
tarrragonparallel-evaluation
Coordinate multi-agent parallel scans and consolidate findings into an action list.
tarrragoncognitive-load-assessment
Quantify cognitive load in software tasks using variable state and call depth metrics.
tarrragonstartup-check
Recover pending handoff tasks and validate Git environment at session startup.
tarrragonmanager
Delegate tasks to multiple agents via tickets and monitor reports.
tarrragonerror-pattern
Query and record error patterns with structured descriptions and preventive measures.
tarrragonsearch-tools-guide
Guide selection of rg, Serena/LSP, and WebSearch for code and documentation retrieval.