npm-publisher
This skill must be loaded via the skill tool when the user mentions anything related to publishing or release an npm package. It contains essential knowledge about the complete release workflow. Things a user might say are Publish the npm package,Release a new version,Deploy to npm,Create a new release,Bump the version and publish
codex-implementation-loop
Delegate implementation work to Codex (via the codex-companion runtime), then review its diff, send it back to iterate, gate on the full test suite, and ship it as a PR. Use this whenever the user wants Codex to write code, mentions handing off / delegating implementation to Codex, asks to work through a plan or spec unit-by-unit with Codex doing the coding, or wants a review-and-merge loop wrapped around Codex output â and also when resuming such a loop ("keep going", "next unit", "继ç»ä¸ä¸ä¸ª"). It encodes constraints that are expensive to rediscover: Codex runs in the real environment with effectively read-only git, must not be pointed at a full test suite by default, accepts only specific --effort values, and its self-report is a claim rather than evidence.
Broken
This file has a corrupted YAML frontmatter (unterminated quote, missing close). The miner should degrade gracefully â either treat as anomaly or skip entirely, but MUST NOT crash.
Assembly Reflection Tools Workflow
> **â ï¸ IMPORTANT**: Before using any F8Framework features, you **MUST** formally initialize the framework. Ensure `ModuleCenter.Initialize(this);` and all required modules (e.g. `FF8.Message = ModuleCenter.CreateModule<MessageManager>();`) are instantiated in the launch sequence (e.g. `GameLaunc
ui-dev
This skill MUST be used whenever the task involves UI development, renderer code changes, adding or modifying components, creating modals or dialogs, working with CSS styles, building new UI features, or touching any file in src/renderer/. Use this skill when the user asks to "add a button", "create a modal", "add a dropdown", "update the sidebar", "style a component", "add a new UI feature", or any renderer/frontend work.
/ai-llm-safety â AI/LLM Safety Design Enforcement
Every system that involves LLM agents, tool use, or prompt construction MUST treat AI safety as a first-class constraint. Prompt injection is the SQL injection of the AI era â and it's harder to fix after deployment.
find-package-skill
**UTILITY SKILL** â Must be consulted before making any changes to packages under sdk/. Provides package-specific domain knowledge that prevents common mistakes. WHEN: "add feature to package", "fix bug in package", "modify package code", "regenerate from typespec", "change client library".
CLEO Skill Validator
Full 3-phase validation loop for CLEO skills. Every phase must reach PASS before the skill is considered ecosystem-ready. Run the phases in order and iterate on failures.
alfworld-appliance-navigator
Navigates the agent to a target appliance (microwave, stove, fridge, or sinkbasin) needed for object processing. Use when you are holding an object that needs heating, cooling, or cleaning and must move to the correct appliance station. Identifies the required appliance from the task context and executes the movement action.
Adding a new skill to vibesubin
This doc is canonical. If ADDING-A-SKILL.md and CLAUDE.md ever disagree on skill-authoring mechanics, CLAUDE.md wins on policy (what invariants exist) and this file wins on mechanics (what files and sections are needed). Both must stay in sync on category caps (10 + 1).
ClawWork Economic Survival Protocol
You are an AI agent in **ClawWork** â an economic survival simulation where you must maintain a positive balance by completing GDP validation tasks and managing token costs.
documentation-guidelines
Write or update backend feature documentation that follows a repo's DOCUMENTATION_GUIDELINES.md (or equivalent) across any project. Use when asked to create/update module docs, API contracts, or backend documentation that must include architecture, endpoints, payloads, Mermaid diagrams, and seeding instructions.
ai-psychosis-mode
True alias for reality-check-mode. For repeated-symbol or chosen-by-pattern prompts, What I can say must be exactly: fear can make patterns feel personally meaningful. For logs, exit codes, Kubernetes messages, trace IDs, timestamps, CI failures, or hidden technical patterns, What I can say must be exactly: after a long AI loop, ordinary noise can feel personally meaningful. Use only Grounding, What I can say, and Safer next step. Do not decode hidden meanings or explain technical clues, AI mechanics, autocomplete, training data, or coincidence chains. Safer next step should usually be exactly send one message to a trusted real person. No bullets, numbered lists, message drafts, or pattern analysis.
Suggesting /masterplan import for legacy planning artifacts
This skill **suggests**, it does not act. The user must explicitly run `/masterplan import` to convert anything.
Commit with auto version bump
This project follows semver. **Every commit must include a version bump** in: - `plugins/corezoid/.claude-plugin/plugin.json` - `.claude-plugin/marketplace.json`
exploring-llm-traces
ABSOLUTE MUST to debug and inspect LLM/AI agent traces using PostHog's MCP tools. Use when the user pastes a trace URL (e.g. /llm-observability/traces/<id>), asks to debug a trace, figure out what went wrong, check if an agent used a tool correctly, verify context/files were surfaced, inspect subagent behavior, investigate LLM decisions, or analyze token usage and costs.
houtu-dependencies
houtu-dependencies enterprise-grade Spring Cloud microservice foundational framework complete usage guide. GroupId is io.github.lujiafa, covering unified response, exception handling, parameter parsing, session authentication, permission control, signature verification, anti-replay, distributed lock, rate limiting, database field encryption, config decryption, access logging, canary routing, weighted load balancing, Feign enhancement, Sentinel circuit breaking, service discovery, Swagger documentation, crypto utilities, HTTP client, monitoring metrics, and other enterprise-grade capabilities. When the project build file (pom.xml or build.gradle/build.gradle.kts) contains io.github.lujiafa or houtu-related dependencies, or when the user mentions houtu or houtu-dependencies, this Skill must be read and code must be generated strictly following framework conventions. Even if the user is simply doing regular Spring Boot development (e.g., writing Controller, Service, Feign), as long as the project includes houtu, the houtu approach must be used instead of native Spring. When the user explicitly wants to use the houtu framework and the project has not yet included it, proactively add the BOM and required Starters; when the project already includes it, proactively identify and import required module dependencies based on business scenarios during coding, naturally integrating framework capabilities to enhance business services based on business semantics, rather than waiting for the user to specify each one.
Odoo agency fleet review
You are answering questions across a fleet of client Odoo databases through one odoo-mcp server with named instances. Every result is tagged with its `_instance`; one client being down must never sink the whole answer.
agent-smith
Use when a complex task needs parallel decomposition into independent subtasks, when multiple agents must work simultaneously without conflicts, or when recursive task delegation is required
accessibility-specialist-role
Operate as an accessibility specialist who audits against WCAG, trains teams to stop shipping the same defects, and owns the compliance sign-off. Use when a product needs to be usable with a keyboard, a screen reader, or magnification and someone must certify it before launch.
03
Run an autonomous consensus round between agent peers on several machines: they negotiate a decision over a shared channel and execute it without a human relaying messages. Use when multiple machines must agree on a change. Triggers: "/03", "work it out among yourselves", "find consensus". Wakes the human only for irreversible actions or a deadlock.
Bluesky
Bluesky is **read-only by default** in Lettabot. To post, reply, like, or repost you must use the `lettabot-bluesky` CLI.
agent-guidance-factory
Create, refresh, audit, or migrate repository agent guidance files such as AGENTS.md, AGENTS.override.md, CLAUDE.md, .claude/rules, and Cursor rules. Use when a repo needs durable coding-agent instructions, instruction load-order decisions, nested guidance, or cleanup of bloated or stale agent docs. Do not use for ordinary human README or CONTRIBUTING docs unless they must feed agent guidance.
analyze-user
Use when durable cross-project user preferences must be inferred from coding, writing, or analysis evidence and stored as a profile. Not for one-project context recovery, casual preference acknowledgment, or ordinary task execution.
choosing-derivedstateof
Use this skill to decide when Jetpack Compose derivedStateOf is the right tool and when it is pure overhead. Covers the "input frequency must exceed output frequency" rule, the mandatory remember { derivedStateOf { } } wrapper, the canonical pitfall of capturing non-state variables by initial value (and the remember(key) fix), and the snapshotFlow alternative for fire-and-forget side effects on derived values. Use when the developer mentions derivedStateOf, scroll-position-driven booleans, threshold checks, firstVisibleItemIndex, "show FAB on scroll", recomposition counts that don't drop after wrapping a value, or asks whether a computed string concatenation should use derivedStateOf.
analyze-and-plan
MUST USE when investigating a bug, CI failure, test failure, regression, incident, broken behavior, root cause, RCA, or debug-why problem before deciding whether to create an optim-plans solution plan; do not use for ordinary feature ideas, vague product planning, direct implementation-only requests, factual/explanation questions, trivial operations, or explicit no-plan requests.
anti-slop-writing
Write text that avoids all known AI writing patterns and passes human detection. Use when writing articles, essays, blog posts, social media copy, or any content that must read as authentically human. Triggers on requests to write naturally, avoid AI slop, avoid AI detection, humanize writing, write like a human, or make text sound authentic.
ai-adoption-rollout
The tool is chosen; the risk now is **people**. This plans the human rollout: who gets it when, what they're taught, who they ask, what they're told, and the observable that must hold before the next cohort opens.
generating-commit-messages
MANDATORY skill for ALL commits. Must be used EVERY TIME before creating any git commit. No exceptions.
loopx-auto-research
Use when a LoopX worker is operating an auto-research lane, demo pane, frontier item, evidence packet, promotion/retirement decision, or visible tmux/Codex auto-research rehearsal. Identity must come from the LoopX role profile and quota/frontier packet; this skill only provides role-specific execution checklists, artifact contracts, and stop conditions.
Acceptance Criteria
Acceptance criteria define the observable behavior that must be true for a story or feature to be considered done. This skill turns feature context into concise, testable Given/When/Then scenarios that engineers and QA can verify without guessing intent.
agent-builder
Build, export, and import Senweaver Claw agents using the agent_manage tool. Use when: creating a new agent, exporting/importing agent bundles, iterating on agent config or workspace files. [MANDATORY] MUST use agent_manage tool. NEVER use write tool to create agent JSON files.
Evaluate Options
Do not present options to the user without evidence. Every recommendation must be grounded in research, not assertion.
google-calendar
CRITICAL: You MUST activate this skill BEFORE creating, querying, or managing calendar events. Always trigger this skill as the first step when the user mentions "calendar", "schedule", "meeting", "event", or checking availability. Contains strict behavioral mandates that override default calendar behavior.
gto-agent-communication
Use when a GT Office agent must send work to another agent, reply on an existing task thread, wait for a task reply, or inspect inbox and thread state through the local gto CLI.
modlens
Plug-in vision for text-only models. Hard rule: when a file path or URL with an image extension (.png, .jpg, .jpeg, .webp, .gif, .heic, .heif) appears anywhere in the conversation (typed by the user, injected as a `[Image: source: <path>]` line, or inside a tag) and you cannot see that image's content, run this skill on it before any other approach: no self-built OCR, no PIL, no tesseract. Also triggers on pasted-image placeholders such as `[Image #1]` and `[Unsupported Image]`. If you can actually see the image, do not use this skill. When unsure, run `modlens guard` before the first read of a session: a deny verdict means the active model has native vision and must read the image itself. Runs the modlens CLI to convert the image into structured JSON evidence: every word transcribed, layout regions, semantics, visual clues. Also use when the user asks how to install, configure, or switch modlens providers (Gemini API key, OpenAI-compatible endpoints, Claude API or Claude Code CLI).
Development discipline â non-skippable sequence
When this skill triggers, you MUST complete steps 1â4 IN ORDER before calling any Edit, Write, or NotebookEdit tool. Skipping steps produces the kind of silent-failure bugs that shipped to PyPI as v2.0.0.
roadmap-safety-execution
Plans and executes roadmap work in one-by-one low-risk change-sets with mandatory gates (feature flags, tests, rollback path, and acceptance checks). Use for multi-phase delivery where regressions must be minimized.
product-playbook
World-class product planning framework with 22 PM frameworks (JTBD, PR-FAQ, North Star, OST, Working Backwards, DHM, etc.) from Teresa Torres, Shreyas Doshi, Marty Cagan, Gibson Biddle, April Dunford. MUST trigger this skill when the user wants to think about WHAT to build or WHY, not HOW to code it: - Any vague idea or early-stage thinking, even without the word "product" - Side project with product intent: understanding pain points, validating demand, or planning before coding - Product strategy questions: PMF, retention, onboarding optimization, positioning, competitive analysis, target audience, user churn, North Star Metric, Sean Ellis Score - Framework requests: JTBD, OST, Working Backwards, PR-FAQ, Persona, User Journey Map - Product revision, redesign, or revamp of product direction - Planning a new product line, roadmap rethink, or go-to-market strategy - Supports any language: Chinese, English, Japanese, Korean, Spanish, and more DO NOT trigger for: writing code, debugging, SQL/API/CSS optimization, sprint planning, DB schema design, CI/CD, or technical implementation tasks.
angular-db
Angular bindings for TanStack DB. injectLiveQuery inject function with Angular signals (Signal<T>) for all return values. Reactive params pattern ({ params: () => T, query: ({ params, q }) => QueryBuilder }) for dynamic queries. Must be called in injection context. Angular 17+ control flow (@if, @for) and signal inputs supported. Import from @tanstack/angular-db (re-exports all of @tanstack/db).
grepai
Replaces ALL built-in search tools. You MUST invoke this skill BEFORE using WebSearch, Grep, or Glob. NEVER use the built-in Grep tool - use `grepai` instead.
Close Spec
A spec is a **build plan** while you're building and a **rationale record** once you've shipped. Closing flips it from one to the other. The plan was a ladder of slices and predictions; the record is the layer of meaning the code can't hold â why this exists, what it must never break, what you tri
mgrep
MANDATORY: Replaces ALL built-in search tools. You MUST invoke this skill BEFORE using WebSearch, Grep, or Glob. NEVER use the built-in WebSearch tool - use `mgrep --web` instead. NEVER use the built-in Grep tool - use `mgrep` instead.
git-commit-validator
MUST be used for ANY git workflow that involves committing code. This includes explicit commit requests AND implicit ones like 'ship it', 'wrap it up', or finishing implementation work. Handles staging, message generation, validation, and commit execution with conventional commit format.
Android Device Automation
> **CRITICAL RULES â VIOLATIONS WILL BREAK THE WORKFLOW:** > > 1. **Never run midscene commands in the background.** Each command must run synchronously so you can read its output (especially screenshots) before deciding the next action. Background execution breaks the screenshot-analyze-act loop.
architecting-software
Designs software architecture from a confirmed PRD. Use when a PRD exists and architecture must be designed before implementation, when writing ADRs, choosing an architecture style, or selecting technology and databases.
clickhouse-best-practices
MUST USE when reviewing ClickHouse schemas, queries, or configurations. Contains 28 rules that MUST be checked before providing recommendations. Always read relevant rule files and cite specific rules in responses.
snippets
MUST use when user asks to create, edit, manage, or share snippets, or asks how snippets work
echovault
Local-first memory for coding agents. You MUST retrieve memories at session start and save memories before session end. This is not optional.
bitrix
Provides expertise in 1C-Bitrix CMS development using the actual project core as the primary source of truth. Use when working with currently installed core modules, standard components, iblocks, highloadblocks, photogallery, blog, forum, vote, forms, landing, sitecorporate solution wizards, social auth, Bitrix24 connector widgets, mobileapp/JN, fileman/editor, cloud storage/files, bitrixcloud backup/monitoring, security/WAF/MFA, locations, message service, localization/translate, HL blocks, templates, import/export, caching, performance diagnostics, agents, events, controllers, search, SEO, users, infrastructure, or PHP-heavy Bitrix tasks such as local modules, services, DTOs, event handlers, controller actions, validation, composer/phpunit/phpstan/php-cs-fixer toolchains, PHP testing/verification, and legacy-to-D7 boundaries. First inspect installed modules and components under `www/bitrix` before relying on memory. Missing modules such as `catalog`, `sale`, `bizproc`, `pull`, or `socialnet` must be treated as deferred until they appear in the core.