Total 44,022 skills
Showing 12 of 44022 skills
Guidance for a developer's first steps on Google Cloud, covering account creation, billing setup, project management, and deploying a first resource.
Install and initialize the correct LaunchDarkly SDK during onboarding by running nested skills in order: detect, plan, apply. Parent onboarding Step 6 is first flag.
Configure the LaunchDarkly hosted MCP server during onboarding. Use when the parent LaunchDarkly onboarding skill reaches Step 4 (MCP). Supports Cursor, Claude Code, Windsurf, GitHub Copilot, and other MCP-compatible agents. OAuth authentication; no API keys for the hosted server.
Create a boolean first flag, add evaluation, toggle on/off for end-to-end proof. Parent onboarding Step 6; uses MCP, API, or ldcli; optional flag-create skill.
Complete reference for writing, running, and iterating on evals (automated conversation tests) for ADK agents. Covers eval file format, all assertion types, CLI usage, and per-primitive testing patterns.
Complete FBA preparation guide. Product labeling, packaging requirements, shipment planning, and compliance with Amazon's fulfillment center requirements. Avoid common rejection reasons.
Ultra-lightweight channel for feature workflows: No need to write design docs, checklists, or conduct phased reviews. Let AI write code directly as it normally would, but before it starts, tell it where the CodeStable knowledge base in the project is and how to search it. This way, the code it writes will have fewer pitfalls and be more consistent with project conventions. Trigger scenarios: Users say "fast mode", "fastforward", "skip all those steps", "just start coding", "help me make xxx" and the requirement is too small to go through the design process.
Discussion entry when ideas are still vague — first conduct triage through 1-2 rounds of dialogue to determine which downstream process this discussion should eventually go to: if the idea is clear enough, proceed directly to feature-design; if the direction of a small requirement is set, continue the discussion within the feature and document it in `{slug}-brainstorm.md`; if a large requirement cannot fit into a single feature, hand it over to roadmap for decomposition. The role of AI is a thinking partner, not a recorder — dig out the real problem the user wants to solve, proactively evaluate when the user brings a solution, and propose alternative directions when necessary. Trigger scenarios: when the user says "I have an idea that's not clear yet", "Let's brainstorm first", "I want to do something but it's still vague", "Let's talk about this area", "The function direction is still undecided", or when the user comes with a specific solution but wants to hear other ideas first. Bugs (go to issue) and refactoring (go to refactor) are not handled here.
Issue Workflow Stage 1 — Convert the user's problem into a reproducible, traceable {slug}-report.md through conversation. The AI only asks "what you saw, how to reproduce it, what should happen" here, and does not guess the root cause for the user (that's Stage 2's responsibility). Meanwhile, this stage is the only official decision point for choosing between the fast track and standard path: Based on the user's description, first review the relevant code; if the root cause can be identified at a glance and the required changes are minor, directly inform the user to take the fast track. Trigger scenarios: The user says "file an issue", "record this bug", "I found a problem". This is the starting point of the issue workflow with no pre-dependencies.
Ultra-lightweight channel for refactor processes - used when changes are clearly too small to go through the full scan → design → apply three-stage workflow. AI directly identifies 1-3 low-risk optimization points, confirms with the user once, modifies in-place using classic methods, and validates itself by running tests. No scan checklist, no design documentation, no multi-step human verification required. Trigger scenarios: User says "quick refactor", "small refactor", "simply optimize XX function", "modify directly", "skip the extra steps", and the scope of changes is clearly localized to a single function / single component with test coverage for self-validation.
When developing new features, follow this sub-process — take the vague idea of "add X capability" through to the acceptance closure, with solution documents archived so that both AI and users can later check the original thinking and decision rationale. Trigger scenarios are focused on adding new capabilities ("develop new feature", "add X", "implement XX"), and do not handle bugs in existing code. This skill only acts as a router, deciding which sub-skill to trigger next among brainstorm / design / fastforward / implement / acceptance based on existing artifacts.
Phase 3 of the issue workflow —— Fix code precisely according to confirmed root causes and solutions, verify the results, and document it in {slug}-fix-note.md. This is the final stage of the issue workflow —— no verification closure means the workflow is incomplete. Two entry points: the standard path is triggered from cs-issue-analyze (with existing {slug}-analysis.md), and the fast track is triggered directly from cs-issue-report (without {slug}-analysis.md, as the root cause was identified by AI through code reading during the report phase). Trigger scenarios: User says "Start fixing the bug", "Fix according to the analysis", "Start modifying the code". During the fix, only modify the files specified in the solution; do not make incidental optimizations or introduce new abstractions —— these actions will cause the scope to expand to an untraceable extent.