Skills

Turns merged work, pull requests, issues, and user-visible behavior changes into accurate, audience-specific release notes with breaking-change and upgrade guidance.

Designs traces, metrics, structured events, privacy controls, dashboards, and alerts for LLM applications so teams can diagnose quality, cost, latency, and tool-use behavior.

Designs and implements secure, ergonomic Model Context Protocol servers with precise tool schemas, transport selection, authentication boundaries, error contracts, and tests.

为 AI 智能体构建可重复的评估，在真实端到端场景中衡量任务成功率、工具使用正确性、事实依据、延迟、成本和安全恢复能力。

Builds regression suites for prompts and agent instructions so teams can detect quality, safety, format, and tool-selection regressions before deploying changes.

Designs reliable JSON and typed outputs for LLM features, including schemas, validation, recovery paths, versioning, examples, and contracts for downstream consumers.

Plans safe database schema and data migrations with backwards compatibility, staged rollout, validation, rollback, performance safeguards, and application coordination.

将陌生的代码仓库整理成简明的工程指南，涵盖架构、约定、本地设置、关键执行路径和低风险的首次贡献。

Defines versioned data contracts between producers and consumers, with ownership, schemas, quality expectations, compatibility rules, and operational change management.

Guides containment, evidence collection, root-cause analysis, customer communication, and remediation when an AI agent takes an unsafe, incorrect, or unauthorized action.

Evaluates retrieval quality for search and RAG systems using grounded test sets, relevance metrics, failure taxonomy, chunking experiments, and actionable remediation.

Designs and implements rate limiting for APIs and services. Recommends an algorithm (token bucket, leaky bucket, fixed or sliding window) for the use case, defines per-key and per-endpoint limits, plans distributed enforcement with Redis, and specifies response headers and 429 handling with retry-after. Produces a design plus reference implementation.

Plans authorized penetration tests for systems you own or are permitted to assess. Defines scope, rules of engagement, and objectives; maps the attack surface; structures phases (recon, mapping, exploitation, post-exploitation, reporting) around a framework like the OWASP Testing Guide or PTES; and specifies safe handling of findings. Emphasizes explicit authorization and non-destructive testing.

Helps turn a game concept into a structured game design document. Captures the core loop, mechanics, progression, economy, controls, level structure, art and audio direction, and target platform and audience. Keeps scope realistic, flags dependencies and risks, and produces a living GDD that a small team can build from.

Turns analysis results into a clear narrative for a specific audience. Selects the key message, orders findings for impact, recommends the right chart for each point, writes plain-language takeaways, and frames actionable recommendations. Helps analysts move from raw numbers to a memo or slide narrative executives can act on.

Produces structured test plans for features and releases. Defines scope and objectives, derives test cases from requirements and acceptance criteria, covers functional, edge, negative, performance, and accessibility cases, sets entry/exit criteria, and maps risk to test priority. Outputs a clear plan with a traceability matrix linking tests to requirements.

Writes conversion-focused ad copy for search, social, and display. Crafts headlines and descriptions within platform character limits, leads with a clear benefit and hook, matches message to audience and funnel stage, and produces multiple variants for A/B testing. Keeps claims honest and includes a strong, specific call to action.

Designs surveys that produce reliable, unbiased data. Turns research goals into clear questions, chooses appropriate scales and response types, avoids leading and double-barreled wording, orders questions to reduce bias, and plans screening and branching logic. Outputs a ready-to-field questionnaire with an analysis plan for each question.

Helps make and document structured decisions. Elicits the options and the criteria that matter, assigns weights, scores each option, computes a weighted result, and runs a quick sensitivity check to show how robust the recommendation is. Produces a clear matrix plus a short written rationale suitable for an ADR or decision log.

Builds discovery and usability interview guides that surface real insight. Translates research questions into open, non-leading prompts, sequences warm-up to deep-dive topics, adds follow-up probes, and applies techniques like the "five whys" and past-behavior questions. Outputs a timed guide with a consent intro and a synthesis template for notes.

Creates accessible color palettes for brands and UIs from a brief or a seed color. Produces primary, secondary, and neutral scales with hex values, suggests semantic tokens (success, warning, error, info), and checks foreground/background pairings against WCAG contrast ratios. Outputs ready-to-use CSS variables or design tokens.