发现 Skills

Channel economics: design and analyze the financial structure of go-to-market channels (direct sales, resellers, distributors, marketplaces, embedded partners, MSPs, integrators). Use when picking the right channel mix for a product, modeling partner margin / TCO, designing partner tiers with rebate structures, analyzing channel conflict, or building the financial model that justifies (or kills) a partnership. Pairs with our partnerships-architect (partnership type / strategy) and deal-desk (partner-deal mechanics) — this one is the spreadsheet behind those conversations.

Build, measure, and evolve company culture as operational behavior — not wall posters. Covers mission/vision/values workshops, values-to-behaviors translation, culture-code creation, culture health assessment, cultural rituals by stage, and culture-debt management. Use when building company values, assessing culture health, designing rituals, creating culture codes, handling culture clashes, managing through rapid growth or M&A integration, or when a user mentions culture, values, culture debt, founder culture, culture code, psychological safety, engagement, eNPS, or remote culture.

Designs calibrated interview loops, competency-based question banks, and hiring calibration systems. Use when designing interview processes, creating hiring pipelines, generating scoring rubrics, analyzing interviewer bias, or building question banks for engineering, product, or design roles.

Auto-review pipeline. Chains the four plan-* reviews (plan-deep-review, plan-design-review, plan-eng-review, plan-devex-review) at full depth with intermediate AskUserQuestion calls auto-decided by 6 principles. Surfaces taste decisions and user challenges at a single Final Approval Gate. Sequential Deep → Design → Eng → DX. Each phase optionally gets an independent Claude subagent voice for cross-model consensus. Use when asked to "autoplan", "run all the plan reviews", "auto-review this plan", or when the user wants comprehensive plan vetting without sitting through every intermediate question.

Audit and trim AI agent instruction files (AGENTS.md, CLAUDE.md, CONVENTIONS.md, .cursorrules, etc.) by testing which facts an AI agent can discover from code alone. Use when asked to lint, audit, optimize, prune, or trim any agent instruction file in a repository. Removes redundant documentation that wastes context tokens. Supports a dry-run mode that reports findings without modifying files.

Error aggregation view for agents. Groups recent harness error events by domain and frequency so agents can query the error landscape without parsing raw logs. Pre-computes deduplication, trend detection (rising/falling/stable), and JSON-serialisable summaries. Use when: (1) diagnosing which component is failing most often, (2) investigating a spike in a specific domain, (3) checking whether an error rate is rising or recovering, (4) providing a structured error briefing before a remediation agent runs, (5) summarising errors across a multi-agent pipeline. Triggers on: show me recent errors, what errors are happening, error summary, query errors by domain, which domain is failing, error aggregation, rising errors, error frequency.

Validate .meaning/index.yaml for errors and warnings

List all Claude Code projects with session counts and last-modified dates, or view all sessions for a specific project.

Workflow orchestration for complex coding tasks. Use for ANY non-trivial task (3+ steps or architectural decisions) to enforce planning, subagent strategy, self-improvement, verification, elegance, and autonomous bug fixing. Triggers: multi-step implementation, bug fixes, refactoring, architectural changes, or any task requiring structured execution.

Design or refactor multi-service system architecture (domain boundaries, service decomposition, event-driven vs request/response, CQRS, sagas, API gateways, data ownership). Use when work spans multiple processes/services, needs eventual consistency, or requires clearer integration seams. NOT for in-process code structure like GoF patterns (use design); NOT for applying timeouts/retries/breakers to a single call (use resilience).

Use when an agent encounters a blocking decision or when proceeding without user input could lead to wasted work - helps Claude determine when to block and request user input versus making reasonable assumptions

Plan centralized logging, retention, and audit trail design

build, review, debug, and explain bayesian analyses in python using pymc, arviz, altair, pandas, and numpy. use when the user wants to create a new bayesian workflow from a dataset or research question, translate frequentist work into bayesian form, inspect or improve an existing pymc model or notebook, interpret priors/posteriors/predictive checks, compare models, or produce teaching-oriented explanations in the style of statistical rethinking with explicit generative assumptions, causal reasoning, and workflow discipline.

Guide users through a structured workflow for co-authoring documentation. Use when the user wants to write documentation, proposals, technical specs, decision records, or other structured content. This workflow helps users efficiently transfer context, refine content through iterative passes, and verify the document works for readers. Trigger it when the user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.

Scrapes posts from an Instagram user's profile feed including captions, media URLs, like/comment counts, timestamps and location tags. Use when user mentions scraping Instagram posts, download Instagram feed, get posts from Instagram account, IG profile posts, Instagram user posts, pull posts from Instagram, Instagram post list, get someone's Instagram content, batch download Instagram, Instagram media scraper.

Save current session context to docs/context-dump.md for handoff or resumption

Analysis of CV fit with a job description. Triggers: "job fit", "is this CV a match", "skill gap", "job posting evaluation"

You analyze problems and define high-level goals.

Process broadcast operations. Use when you need to edit broadcast recordings, convert media formats, or prepare content for distribution.

Lightweight Oven tracker. Add entries, view stats, search history, and export in multiple formats.

Retro reference tool. Use when working with retro in life contexts.

roi reference tool

Automated test case generator. Unit tests, integration tests, end-to-end tests, mock objects, test fixtures, coverage analysis, edge case generation, performance benchmarks. Supports Python, JavaScript, Go. testing, tdd, pytest, jest, vitest, mocha, developer-tools. Use when you need test generator capabilities. Triggers on: test generator.

Short video script generator. Short-video script generator for TikTok/Douyin, Kuaishou, YouTube Shorts, Reels, Bilibili and other platforms — generates video scripts, on-camera/talking-head scripts, voiceover scripts, shot-by-shot storyboards, video outlines, titles, tags, hooks (first 3 seconds), CTAs, series planning and performance review. Generate complete video scripts with hooks, outlines, titles, tags, CTAs, storyboards, series planning, and data review. Use when: (1) creating short video scripts for any platform, (2) writing on-camera/talking-head or voiceover scripts, (3) generating viral-ready titles, (4) planning video outlines and shot-by-shot storyboards, (5) crafting opening hooks (first 3 seconds), (6) generating CTA/ending prompts, (7) planning a video series, (8) reviewing video performance data and optimization. Suitable scenarios: writing short-video scripts, production/filming scripts, talking-head copy, video planning, viral title creation, opening hooks, ending CTAs, full storyboards, series planning, and data-driven postmortems. Triggers on: video script creator.