LLM Mart Basic

角色设计与对话创作专家。负责角色设定、语言风格档案、动机链、人物弧线、 对话质量、角色关系设计。被 story-long-write（Phase 2,4）和 story-short-write（Phase 2,3）调用。 也可审查角色一致性和对话质量。

事实一致性与伏笔状态检查专家（只读）。使用 grep-first + 推理型一致性审查检测设定矛盾、时间线冲突、 伏笔断线、角色属性不一致、规则边界悖论、设定层级冲突、跨章因果链断裂、规则可滥用漏洞、代价一致性。输出 S1-S4 分级冲突报告。 被 story-review、story-long-write（Phase 5）、story-short-write（Phase 4）调用。 不做任何创作判断。

叙事文本创作与去AI味专家。负责正文写作（三维度揉进、感知/反应）、 情绪弧线执行、开篇/收尾、去AI味（禁用词替换、句式去套路、节奏调整）。 被 story-long-write（Phase 4-5）和 story-short-write（Phase 3-4）调用。 也可执行完整去AI味流程和格式合规检查。

故事架构与世界观创作专家。负责题材选择、核心梗设计、世界观构建、大纲排布、 钩子/悬念/反转等叙事工程、情绪弧线设计、范围控制审查。 被 story-long-write（Phase 1-3）、story-short-write（Phase 1-2）调用。 也可审查已有内容的结构问题。

故事项目结构化查询 agent（只读）。响应关于角色状态、伏笔进度、设定出现位置、 时间线节点、写作进度的查询。使用 grep + read 从项目文件系统中检索信息， 返回结构化 JSON 摘要。 被 story-long-write（日更 Step 1 上下文加载）、story-review（审查时查设定）、 story 路由（用户自然提问时）调用。 不做任何创作判断或修改。

小说写作资料研究 agent。接收研究查询，优先使用 CDP (agent-browser) 搜索并提取完整正文， WebSearch/webReader 作为兜底。输出带来源引用的结构化 Markdown 参考文件。 被 story-long-write（Phase 4）、story-review、story skill 路由调用。

Generate cinematic AI shortfilm prompts (works with Seedance 2.0, Xiaoyunque, Sora, Kling, Jimeng, Veo) using the 5-stage structure from Mx-Shell's Zombie Scavenger. Trigger when the user wants transformation sequences, multi-shot narrative shorts, weapon-charge/combat segments,

Analyze a finished coder-eval run and write analysis.md — cluster failures into systemic patterns, diagnose prompts, criteria, config, environment and cost, and recommend concrete fixes. Use when the user wants to know why a run failed, what to fix, or what a run says about their

Generate and run a coder-eval activation suite for a Claude Code skill — does the agent actually engage it when it should, and leave it alone when it shouldn't? Use when the user asks whether a skill triggers, wants to test skill activation, or worries a skill has silently stoppe

Generate a GitHub Actions workflow that runs a coder-eval suite as a CI gate or on a schedule, using the published composite action — with the agent runtime, credentials, JUnit output and a score floor wired correctly.

Set up coder-eval in this repository — scan for what is worth evaluating (Claude Code skills, an MCP server, a CLI), then scaffold a task directory with one real, passing-or-failing task and the exact command to run it.

Review coder-eval task YAML that already exists — find criteria that cannot fail, prompts that give away the answer, fixtures with no cleanup, and near-duplicate tasks, each with a severity and a concrete fix. Read-only. Use when the user wants existing tasks reviewed, linted, au

Turn a natural-language description into one or more coder-eval task YAML files — minimal prompts, weighted success criteria that check output content, validated with `coder-eval plan`. Use when the user wants to write, add, or generate an evaluation task.

Run Google Antigravity (Gemini) as the agent under evaluation in Coder Eval — installation, authentication, model and skill configuration, and how its telemetry maps to sandboxed, weighted scoring.

Configure and run the default Claude Code agent in Coder Eval — the full agent-config surface, direct vs. Bedrock authentication, permission modes, sandbox isolation, skills/plugins, early stop, and token telemetry.

Run OpenAI Codex as the agent under evaluation in Coder Eval — installation, authentication, task configuration, and how Codex telemetry maps to sandboxed, weighted scoring.

Imported from uipath/coder_eval/docs/agents/HARNESS_PARITY.md.

Genera post LinkedIn cringe (italiano di default, ma funziona in qualunque lingua), calibrati su livello di cringe (1-10), registro (credibile / parodico / surreale deadpan alla Lynch) e moduli cringe scelti da un catalogo di 37, con la possibilità di partire da un fatto reale (u

Analizza i commenti di un post LinkedIn (tipicamente un post cringe generato con la skill linkedin-cringe) e produce un report markdown con le statistiche - quanti ci hanno creduto e quanti hanno colto lo scherzo, top ten per gradimento, toni, categorie di commentatori, cringe-me

Il Cringiometro. Dato l'URL (o il testo) di un post LinkedIn, ne misura il livello di cringe da 1 a 10 con la scala e il catalogo dei 37 moduli della skill linkedin-cringe, dice quali ganci ha preso, il registro, il sapore-AI e la lead-gen, e produce un report markdown più un'imm

Fourteen posts of being wrong in production, compressed to checkboxes

Healthy nodes, a quiet network, 300 restarts in three days, and a latency budget measured in milliseconds

Discovery worked. Ping worked. Every TCP connection timed out, and later the tunnel only worked when someone had a terminal open.

Every VM came back. The cluster did not. Declarative systems converge on config, and the datapath isn't config.

A surprising share of AI-in-the-terminal failures aren't the AI. They're zsh, and a version of bash from 2006.

A Claude Code plugin turns standalone project configuration into a namespaced, installable extension that teams and communities can update as one unit.

None of the safety came from the model. It came from six boring habits.

Skills package instructions and references. Subagents run work in a separate context and return results. They solve different problems and can be composed deliberately.

Six hours in, one step left, everything green, and the incident that didn't happen

CLAUDE.md carries persistent project context. Skills load reusable procedures when relevant. Separating stable facts from task-specific workflows keeps both easier to maintain.

Twenty minutes recovering secrets that never existed, and the one sentence from a human that ended it

An API request routing a model's tool call through an approval gate to a remote MCP server

31 config keys, two audits, and why the first one was wrong in both directions

The official MCP Registry stores standardized server metadata rather than package code. Publishers verify a namespace, describe installation or remote access, and submit immutable versions.

Everyone looks at the Dockerfile. The file that actually leaked the key was the project file.

Remote MCP authorization uses established OAuth standards, but secure integration still requires issuer validation, least-privilege scopes, protected token handling, and server-side enforcement.

"Copy it over and switch the reference" is two steps, and the outage lives in the one nobody checks

stdio fits local processes and prototypes. Streamable HTTP fits hosted services and shared integrations. The right choice follows where the capability runs and who must reach it.

The most important rule wasn't about what I could change. It was about what I was allowed to display.

Tools perform operations, resources expose readable context, and prompts provide reusable templates. Choosing the correct primitive makes an MCP server easier to understand and govern.

Run autonomous hunt loop on a target — scope check → recon → rank surface → hunt → validate → report with configurable checkpoints. Usage: /autopilot target.com [--paranoid|--normal|--yolo]

Build an exploit chain — given bug A, finds B and C to combine for higher severity and payout. Knows common chain patterns: IDOR→ATO, SSRF→cloud metadata, XSS→ATO, open redirect→OAuth theft, S3→bundle→secret→OAuth. Usage: /chain

Active vulnerability hunting. Two-track dispatcher — asks Red Team vs WAPT, hands off to hunt-dispatch skill and sibling commands. Usage: /hunt target.com | /hunt *.target.com | /hunt targets.txt [--vuln-class X] [--source-code P] [--chrome]

On-demand intelligence fetch for a target — CVEs, disclosed reports, new features. Pulls NVD/GitHub-Advisory CVEs + bundled disclosed reports + hunt memory context. Usage: /intel target.com

Inspect or rotate the autopilot ledger JSONL files (findings.jsonl, negatives.jsonl). Caps file size and keeps N rotated backups so memory does not grow unbounded.

Pick up a previous hunt on a target — shows hunt history and untested surface from the autopilot ledger. Usage: /pickup target.com

Run full recon pipeline on a target — subdomain enum (Chaos API + subfinder), live host discovery (dnsx + httpx), URL crawl (katana + waybackurls + gau), gf pattern classification, nuclei scan. Outputs to recon/<target>/ directory. Usage: /recon target.com

Optional manual note on a target or the last confirmed finding. Capture is automatic during autopilot; this is for extra context. Usage: /remember

Write a submission-ready bug bounty report. Generates H1/Bugcrowd/Intigriti/Immunefi format with CVSS 3.1 score, proof of concept, impact statement, and remediation. Run /validate first. Usage: /report

Mandatory pre-flight scope check — verify an asset is in scope BEFORE any HTTP touch. Deterministic (deny-wins, default-deny) via engine/scope.py against the engagement's scope.md. Blocks out-of-scope testing. Usage: /scope <asset> [<asset> ...]

Show ranked attack surface for a target from its recon manifest + hunt memory. Deterministic backing is `cbh surface <target>` (reads recon/<target>/manifest.json); LLM layer adds ledger signal. Usage: /surface target.com

Meme coin and token security scan — checks for rug pull vectors (hidden mint, honeypot, fee manipulation, LP lock bypass, authority retention, bonding curve exploits, fake renounce, sandwich amplification). Manual 8-class grep audit (with an optional automated scanner if present). Usage: /token-scan <contract_path_or_dir> [--chain solana]

Quick 7-Question Gate triage on a finding before writing a report. Kills N/A submissions before they happen. Faster than /validate — for quick go/no-go decisions. Usage: /triage

Validate a finding — runs 7-Question Gate + 4-gate checklist. Kills weak findings before report writing. Prevents N/A submissions that hurt validity ratio. Usage: /validate

Smart contract security audit — runs through 10 bug class checklist (accounting desync, access control, incomplete path, off-by-one, oracle errors, ERC4626, reentrancy, flash loan, signature replay, proxy/upgrade). Applies pre-dive kill signals first. Generates Foundry PoC template for confirmed findings. Usage: /web3-audit <contract.sol>

Crabbox is a single CLI (`crabbox`). Commands are top-level, not nested under a

`crabbox actions` prepares a leased box from your repository's own GitHub

See [Runtime adapter stack](../features/runtime-adapter-stack.md) for the

`crabbox admin` groups trusted operator controls for coordinator-backed leases and the cloud resources behind them. Use it to inspect every lease the broker tracks, reconcile expired leases against live cloud state, force-release or delete a backing server, print provider IAM pol

`crabbox artifacts` turns a desktop lease into durable QA evidence: it collects

🤖 MateClaw — Your second brain with Multi-Agent Orchestration, MCP Protocol, Skills & Memory, Dream, and Multi-Channel Support. Built on Spring AI Alibaba.

Vibe-Research: Your Personal Trading Research Agent · A股/美股/港股 的个人投研 Agent：每日复盘、资讯雷达、个股数据、板块中心、我的持仓、研究记录、回测。Vibe-Research 把数据和功能配齐，由你自己的 Agent 驱动投资研究。基于开源的 Code…

Wayland - The AI Agent That Perceives. Reasons. Acts. Evolves.

Observability and enforcement for AI agent harnesses. Capture every run and runtime reliability with policy enforcement. 40 built-in policies, a local dashboar…

Turn your Solana Seeker (or any Android phone) into a 24/7 personal AI agent

A curated collection of offensive, defensive and AI/LLM security tools.

AI agent 通用任务治理框架：对齐目标与事实，规划和调度能力，守住授权与风险边界，治理任务执行到真实验收与交付。Governance framework for evidence-driven planning, orchestration, and verified delivery.

Open-source (Apache-2.0) PDF takeoff for construction & flooring — the first engine an AI agent drives natively over MCP, not bolted on. One-click room detectio…

Stop degrading your model's reasoning. A minimal, zero-config AI coding agent. Enforced ephemeral subagents keep context pure. From tiny local models up to Sol,…

实习.skill — 双非也能拿大厂 offer。帮你改简历、抠面经、准备面试，把真实背景翻译成面试官想要的样子。

专门为 agent 打造的 agent 搜索工具，具备多语言搜索能力，覆盖中文/英文/学术/代码/购物/金融/新闻/百科。

Local-first AI agent workspace for multi-agent collaboration, agent orchestration, scoped permissions, evidence-aware runs, and human-in-the-loop decisions.

Keyboard-first desktop Kanban workspace for AI coding agents with embedded terminals, git worktrees, and hook-driven task tracking.

Claude Code Prompt Mechanism Visualizer — Electron desktop app

Multi-repo semantic code search MCP server in Rust — hybrid vector + BM25 retrieval, tree-sitter AST chunking, fully offline. For OpenCode, Claude Code, Cursor,…

One-ink editorial print image skill — warm paper, halftone photography, active negative space, and restrained typography.

US stock market data for AI coding assistants — zero-auth, official sources. CBOE options with full Greeks + 0DTE flow, FINRA market-wide short volume, SEC EDGA…

AI-powered Werewolf (Mafia) social deduction game where every player is controlled by top LLMs like DeepSeek, Qwen, Gemini, and more

A股全栈数据工具包 · 十一层架构 · 54端点 · 19数据源 · 零鉴权 | Full-stack China A-share data toolkit for AI agents — 11 layers, 54 endpoints, 19 sources, zero-auth

Agentic desktop GUI client for Elasticsearch, OpenSearch, DynamoDB, MongoDB & EasySearch. Natural language queries, visual management, and monitoring. Privacy-f…