Coding standards, frameworks, refactoring, and code review (18.6K skills)
Submit a new Wall of Apps entry or update an existing one using the `asc apps wall submit` CLI flow. Preview the proposed change first, then review the pull request plan and the resulting entry update.
Bulk-create or update localized display names for subscriptions, subscription groups, and in-app purchases across App Store locales. It can also handle descriptions where supported, using version-scoped asc commands and checking existing entries before writing changes.
Automate iOS simulator screenshot workflows: build and launch an app, capture screens through AXe, add device frames, review the resulting images, and upload them with asc. Useful for preparing localized screenshot sets as well as single captures.
Turn a codebase into Mermaid diagrams such as architecture, ER, sequence, class, and flow diagrams. It analyzes relevant files, writes diagram source, renders SVG and PNG images, and checks the rendered result for readability and correctness.
Plan and build MCP servers that let AI agents work with external APIs through tools. The guide covers research, implementation, testing, and registration, with patterns for Python FastMCP and the Node/TypeScript SDK, including schemas, pagination, authentication, and useful error messages.
Revise an existing agent skill in response to feedback. The skill guides the agent to inspect the skill files, plan focused changes, update instructions or supporting scripts, and validate the result before packaging it.
Review JSON execution traces and user feedback to find where a skill fails, takes too many steps, or uses too many tokens. Identify likely causes and apply focused improvements to the skill’s existing instructions or scripts.
Investigate an agent run by reviewing its execution trace, a record of steps and tool calls. Start with a run overview, then inspect relevant inputs, outputs, errors, and token use to explain what happened or diagnose a failure.
Build and launch an MVP, from product scope and landing page through implementation, payment integration, deployment, and quality checks. It also outlines post-launch monitoring, including checks for uptime, core flows, payments, errors, and performance.
Maintain documentation that AI agents can discover and read. This skill explains the requirements for `llms.txt` indexes, page Markdown, skill and MCP discovery endpoints, and HTML-to-Markdown parity, with guidance for keeping these checks intact as docs change.
Capture before-and-after screenshots of changed Sentry UI in the real app, organize them by page and state, and attach them to a pull request description. The skill also checks image differences and removes sensitive values before upload.
Work through a pull request’s actionable CI failures and high or medium priority review feedback. Diagnose root causes, make and verify fixes, then monitor checks and new feedback in a loop, stopping when only human approval or merge gates remain.
Create or update concise AGENTS.md and CLAUDE.md files that give coding agents actionable project guidance. Inspect existing commands, conventions, and documentation, then point to authoritative references instead of duplicating them.
Create a Git branch using Sentry naming conventions. Determine the work type and a short ASCII, kebab-case description from the provided request or local changes, choose a base branch using repository information, avoid name collisions, and report the branch created.
Document Sentry API endpoints with OpenAPI schemas, typed responses, and request parameters. Compare declared types with runtime responses, reuse canonical response types, migrate legacy path docs, and validate the generated specification.
Refine recently changed code to make it clearer and easier to maintain while preserving its behavior. The skill follows project-specific conventions, removes unnecessary complexity, and favors readable, explicit code over compact tricks.
Remove assigned Python modules from mypy exclusion settings in pyproject.toml, fix typing issues within the agreed scope, and run the requested checks. Return a structured batch summary covering changed modules, fixes, validation, and remaining blockers.
Add a GitHub issue reference and, when available, its linked Linear ticket to the current pull request description. The skill checks the issue’s linkback comment, preserves the existing description, and replaces an existing Issues section instead of duplicating it.
Review pull requests and code changes using Sentry engineering practices. Check for runtime errors, performance problems, security risks, design mismatches, and test gaps, then give polite, actionable feedback focused on reducing risk rather than enforcing personal style preferences.
Inspect a repository’s languages, frameworks, build tools, services, and existing Claude Code settings. Generate a project-specific `settings.json` permissions allowlist containing read-only commands, with MCP suggestions when relevant.
Create or improve agent skills by gathering and synthesizing source material, writing skill files, refining their descriptions, and handling registration and validation. The workflow also supports updates based on positive or negative examples and maintenance of skill documentation.
Draft commit messages or guide commits using Sentry-style Conventional Commits. Choose an allowed change type, write a concise subject and useful body, and add issue references or breaking-change notes while avoiding personal or sensitive information.
Prepare or refresh a pull request title and reviewer-facing description based on the full branch diff. Explain the change and its effects in a concise form, adding migration, compatibility, risk, or review guidance when it helps reviewers assess the work.
Review Django code for verified performance problems such as repeated database queries, unbounded querysets, missing indexes, and write loops. Trace how data is used and confirm scale and impact before reporting findings with evidence and suggested fixes.