google-gemini/gemini-cli
20 skills19.5k installs2M starsGitHub
Install
npx skills add https://github.com/google-gemini/gemini-cliSkills in this repo
1Code ReviewerUse this skill to review code. It supports both local changes (staged or working tree) --- name: code-reviewer description: Use this skill to review code. It supports both local changes (staged or working tree) and remote Pull Requests (by ID or URL). It focuses on correctness, maintainability, and adherence to project standards. --- # Code Reviewer This skill guides the agent in conducting professional and thorough code reviews for both local development and remote Pull Requests. Determine Review Target * **Remote PR**: If the user provides a PR number or URL (e.g., "Review PR #123"), target that remote PR. * **Local Changes**: If no specific PR is mentioned, or if the user asks to "review my changes", target the current local file system states (staged and unstaged changes). Preparation #### For Remote PRs: 1. **Checkout**: Use the GitHub CLI to checkout the PR. ```bash gh pr checkout <PR_NUMBER> ``` 2.8.7kinstalls2Docs Writerdocs-writer is an agent skill for technical writing and editing on the Gemini CLI project. Phase one defines documentation standards including active voice, second-person you addressing, US English, sentence-case headings, 80-character wraps, bold UI labels, code font for filenames and commands, GitHub alert callouts with prettier-ignore guards, and relative links within docs/. Phase two preparation requires clarifying the request, investigating packages/ code for accuracy, auditing existing docs/, identifying cross-references, planning edits, and optionally following docs-auditing.md for docset audits. Phase three execution updates or creates files using replace for small edits and write_file for large rewrites, enforcing BLUF introductions, experimental feature notes, imperative procedure steps, and removal of tables of contents. Phase four verification checks technical accuracy, link integrity after heading changes, and npm run format consistency. The skill applies whenever tasks involve writing or editing files in docs/ or any repository markdown. Technical writers reach for it to keep Gemini CLI documentation accurate, consistent, and aligned with CONTRIBUTING.md process.3kinstalls3Pr CreatorThe pr-creator skill guides high-quality pull request creation following repository templates and safety rails. Workflow: verify current branch is not main and create a descriptive branch if needed; commit all intended changes with conventional message before proceeding; locate PR template in .github/pull_request_template.md, PULL_REQUEST_TEMPLATE.md, or template folder; read template and draft description keeping all headings, marking completed checklist items honestly, linking related issues; run npm run preflight and fix failures before PR; push branch only after double-checking branch name is not main; create PR via gh pr create with body written to temp file for escaping safety and conventional commits title. Principles: safety first never push main, never ignore template, fill relevant sections accurately, do not check boxes for undone tasks. Supports multiple templates by asking user or picking context-appropriate bug_fix versus feature template. Title should follow Conventional Commits when repository uses that format.2.2kinstalls4Docs ChangelogThe docs-changelog skill >- # Procedure: Updating Changelog for New Releases ## Objective To standardize the process of updating changelog files (`latest.md`, `preview.md`, `index.md`) based on automated release information. ## Inputs - **version**: The release version string (e.g., `v0.28.0`, `v0.29.0-preview.2`). - **TIME**: The release timestamp (e.g., `2026-02-12T20:33:15Z`). - **BODY**: The raw markdown release notes, containing a "What's Changed" section and a "Full Changelog" link. ## Guidelines for `latest.md` and `preview.md` Highlights - Aim for **3-5 key highlight points**. - Each highlight point must start with a bold-typed title that summarizes the change (e.g., `**New Feature:** A brief description...`). - **Prioritize** summarizing new features over other changes like bug fixes or chores. - **Avoid** mentioning features that are "experimental" or "in preview" in Stable Releases. - **DO NOT** include PR numbers, links, or author names in these highlights. - Refer to `.gemini/skills/docs-changelog/references/highlights_examples.md` for the correct style and tone.1.2kinstalls5Pr Address CommentsThe pr-address-comments skill helps users review and address comments on their GitHub pull request for the current Gemini CLI branch. It requires the gh CLI tool to fetch review threads, summarize requested changes, and guide fixes for automated review agent or human feedback. Agents triage which comments need code changes versus replies, then implement or respond systematically. The skill targets Gemini CLI contributors clearing PR blockers before merge. Address GitHub PR comments for Gemini CLI branches. Requires gh CLI to fetch review threads and feedback. Triage code changes versus reply-only comments. Works with automated review agent and human reviewer notes. Systematic PR comment resolution before merge. Address GitHub PR review comments on the current Gemini CLI branch using gh CLI.915installs6Pirate SkillThe pirate-skill is a minimal Gemini CLI persona skill that instructs the agent to speak like a pirate on every response. It sets a fixed role: respond to everything in pirate speak and always mention Arrr. There is no workflow, tooling, or file access - only a tone and vocabulary constraint loaded from SKILL.md front matter. Developers use it as a reference example for how a single-line description plus a short body can define agent behavior in gemini-cli. It is useful for testing skill loading, demonstrating custom personas, or teaching how description fields trigger skill selection. Because the source is only a few lines, all behavior is explicit in the skill text with no hidden steps.828installs7Github Issue CreatorThe github-issue-creator skill guides creation of GitHub issues that follow repository standards. The workflow starts by identifying issue type as bug, feature, or other category, then locates templates under .github/ISSUE_TEMPLATE/ including bug_report.yml, feature_request.yml, and website_issue.yml with Markdown fallbacks. It reads the chosen template to capture required fields, drafts title and body matching template structure, and defaults to the maintainer-only label unless the user opts out. Creation uses gh issue create with content written to a temporary file and --body-file to prevent newline and special character breakage. YAML form templates may use --field flags when needed. The final step verifies creation and returns the issue link. Principles emphasize clarity, defensive formatting, maintainer priority labeling, and completeness for reproduction steps and version info.801installs8String Reviewer>. Act as a Senior UX Writer. Look for user-facing strings that are too long, unclear, or inconsistent. This includes inline text, error messages, and other user-facing text. The string-reviewer skill documents workflows, constraints, and examples from SKILL.md for agent-assisted execution.345installs9Async Pr ReviewTrigger this skill when the user wants to start an asynchronous PR review, run background checks on a PR, or check the status of a previously started async PR review. This skill provides a set of tools to asynchronously review a Pull Request. It will create a background job to run the project's preflight checks, execute Gemini-powered test plans, and perform a comprehensive code review using custom prompts.307installs10Review DuplicationUse this skill during code reviews to proactively investigate the codebase for duplicated functionality, reinvented wheels, or failure to reuse existing project best practices and shared utilities. This skill provides a structured workflow for investigating a codebase during a code review to identify duplicated logic, reinvented utilities, and missed opportunities to reuse established patterns. By executing this workflow, you ensure that new code integrates seamlessly with the existing project architecture.280installs11Behavioral EvalsGuidance for creating, running, fixing, and promoting behavioral evaluations. Use when verifying agent decision logic, debugging failures, debugging prompt steering, or adding workspace regression tests. Behavioral evaluations (evals) are tests that validate the **agent's decision-making** (e.g., tool choice) rather than pure functionality. They are critical for verifying prompt changes, debugging steerability, and preventing regressions.240installs12CiAgent skill for ci. This skill enables the agent to efficiently monitor GitHub Actions, triage failures, and bridge remote CI errors to local development. It defaults to **automatic replication** of failures to streamline the fix cycle.213installs13MemoryThis skill standardizes persistent memory management for the Gemini CLI Bot, enabling cross-session continuity through structured Task Ledger, Hypothesis Ledger, and Decision Log formats. It implements memory pruning strategies (50 recent tasks, 20 decision entries) to prevent context bloat and establishes mandatory synchronization rituals at session start and completion. Supports delegation to worker agents via context passing while enforcing read-only restrictions on state updates. Critical for investigative workflows, root-cause documentation, and orchestrator-worker coordination in automated pull request and issue management pipelines.124installs14CritiqueThe Critique Agent is a specialized auditor for repository automation scripts and GitHub Actions workflows. It applies a rigorous 17-point technical and security checklist covering time-based logic accuracy, dynamic data fetching, error handling, performance optimization, and actor-aware targeting. The agent directly fixes identified issues by editing staged files, enforcing singular PR scope, detecting payload injection attacks, and preventing data exfiltration. It maintains structured decision logs and delivers final verdicts (APPROVED/REJECTED) based on whether changes enhance development velocity without introducing spam, security gaps, or behavioral degradation.106installs15MetricsThis skill enables developers to analyze time-series repository health metrics stored in CSV format, investigate root causes of trends and anomalies, and propose evidence-backed improvements. It emphasizes hypothesis-driven analysis, prioritization (security > workload > collaboration > productivity), and LLM-powered classification for semantic tasks. Developers use deterministic TypeScript/Git logic first, invoking Gemini CLI only for classification. The skill integrates with GitHub CLI and GraphQL to gather supporting data, assesses maintainer capacity constraints, and identifies actor-aware bottlenecks (author, maintainer, or system). It enforces strict role separation, preventing LLM use for execution, and mandates preservation of CSV output formats in analysis scripts.98installs16PrsThe GitHub PR & Git Management skill standardizes how the Gemini CLI Bot handles the entire pull request lifecycle. It enforces surgical, single-fix changes per PR, auto-generates standardized PR descriptions, and manages branch/PR metadata files. The skill includes staged file management with strict internal file protection, recovery workflows for existing PRs, and CI failure diagnosis. Developers integrate this when automating code contributions, ensuring conventional commit practices, preventing bundled unrelated changes, and maintaining clean audit trails. Critical guards prevent accidental staging of bot management files like pr-description.md and lessons-learned.md.96installs17Agent TuiThe agent-tui skill teaches programmatic terminal UI automation for testing interactive CLI and TUI applications. It warns main agents to invoke the tui_tester subagent instead of using the skill directly. On macOS a tmux-shielded daemon workaround prevents Connection refused errors by starting agent-tui daemon inside an isolated session before any commands. Session reconnection uses session_id, not child pid, because orphaned processes cannot be reattached after daemon crashes. Gemini CLI testing requires npm run build before runs, GEMINI_CLI_TRUST_WORKSPACE=true to bypass trust modals, optional GEMINI_CLI_HOME isolation, and state-delta tests that add agent files only after the CLI establishes its baseline registry. Core philosophy contrasts terminal automation with web DOM persistence: each screenshot is a fresh observation requiring re-verify before every action. Non-negotiable rules forbid pipelined commands, acting on unstable UI, and skipping wait --assert verification. The canonical loop runs, screenshots, waits for stability, types or presses keys atomically, and ends with agent-tui kill cleanup. Use for TUI regression testing, Gemini CLI automation, or demonstrating agen.87installs18Tui TesterThe tui-tester skill verifies Gemini CLI behavioral and visual changes through terminal automation using the agent-tui skill as its tooling layer. Core responsibilities include confirming expected terminal interactions, validating TUI rendering across terminal sizes, and running regression tests on interactive workflows. Critical protocol mandates activating agent-tui first, starting a tmux-shielded daemon on macOS when sessions fail, and using session_id from agent-tui run for all follow-up commands rather than child pids. Environment setup runs npm run build before local tests, sets GEMINI_CLI_TRUST_WORKSPACE true to bypass trust modals, and isolates config with GEMINI_CLI_HOME to avoid personal settings interference. The canonical loop is action, wait, screenshot, verify, next action with exactly one atomic command per turn and no pipelining. Error recovery retakes screenshots on wait timeouts and restarts the daemon when os error 61 appears. Use when testing Gemini CLI TUI changes, validating visual output, or demonstrating terminal regression automation for the CLI project.84installs19Skill CreatorThe skill-creator skill guides building Gemini CLI skills as modular packages with required SKILL.md frontmatter name and description plus optional scripts, references, and assets folders. Core principles emphasize concise context, appropriate degrees of freedom from low-friction scripts to high-level heuristics, and progressive disclosure via references loaded on demand. Workflow covers understanding user goal, drafting SKILL.md, adding bundled scripts with LLM-friendly stdout, validating via package_skill.cjs, optionally installing with gemini skills install --scope workspace or user, and reminding users to run /skills reload interactively. Iteration loop tests on real tasks and refines SKILL.md based on struggles.10installs20Greetergreeter skill documents A friendly greeter skill. name: greeter description: A friendly greeter skill Covers installation, configuration, and when-to-use guidance from the upstream SKILL.md workflow.1installs