Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/neopen/video-clip-agent/claude-mdgit clone --depth 1 https://github.com/neopen/video-clip-agentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/neopen/video-clip-agent/claude-md)<a href="https://agentmods.dev/instructions/neopen/video-clip-agent/claude-md"><img src="https://agentmods.dev/badge/instructions/neopen/video-clip-agent/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01211 | $0.01211 |
| Opus 5 | $0.00606 | $0.00606 |
| Sonnet 5 | $0.00242 | $0.00242 |
| Haiku 4.5 | $0.00121 | $0.00121 |
Grade A, and why
video-clip-agent CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PenClip — 视频混剪智能体
基于星型中枢分发架构的 AI 视频混剪系统：自然语言驱动视频分析、片段匹配与自动合成。核心理念「人机协同」—— AI 辅助而非替代创作者。
当前状态
- 版本：V0.1（DDD 分层骨架），开发分支
dev - 顶层架构：星型中枢分发（
CentralHub）+ 线性链路工作流（首个注册能力） - 意图识别：V0.1 关键词匹配（
IntentRecognizer），V0.2 才升级 LLM 语义理解 - 图编排：
LangGraphEngineV0.1 仅 stub，V0.2 真正接入 LangGraph
分层架构（依赖方向：外层 → 内层，内层禁止反向依赖）
domain/ 领域层：实体(Pydantic) + 值对象(dataclass/enum) + 仓储接口 —— 零依赖
core/ 应用层：hub 分发 / orchestration / state / event —— 依赖 domain
agents/ 智能体：BaseAgent + Planner/Analyzer/Matcher/Composer —— 依赖 domain+core
services/ 服务层：llm/clip/ffmpeg/scene_detect/vector/file —— 多为 V0.2 stub
infrastructure/ 基础设施：postgres/minio/redis/cache/messaging —— 实现仓储接口，多为 stub
api/ 接入层：v1 REST / middleware / websocket / schemas
cli/ CLI 入口
plugins/ 插件系统 + Docker 沙箱（V2.0）
遗留层（仍在使用，逐步迁移）：client/（多 AI 提供商）、tools/（6 个视频分析工具）、utils/（ffmpeg/redis/path/log 等）、config/config.py、logger.py
Bridge 层（向后兼容转发，勿在此改逻辑）
顶层 hub/、graph/、state/ 是兼容转发模块，真实实现在 core/、domain/：
penclip.hub.hub_core→penclip.core.hub.central_hub（get_hub/CentralHub/HubRequest/HubResponse）penclip.graph.hub_graph→penclip.core.orchestration.graph_engine（get_graph/LangGraphEngine）penclip.state.models→IntentType(domain) +TaskLifecycleStage(core.state)
关键入口
from penclip.hub.hub_core import get_hub # 中枢单例
from penclip.graph.hub_graph import get_graph # 图引擎（stub）
from penclip.state.models import IntentType, TaskLifecycleStage
from penclip.domain.value_objects.intent import IntentType # 真实定义处
启动：python main.py（→ penclip.app:app FastAPI，默认 http://localhost:8000，`/docs` 交互文档）
核心契约（开发必读）
IntentType（str Enum）：30+ 意图，6 类 —— 规划/分析/素材/效果/状态/通用CapabilityDeclaration（dataclass）：name+intents+input_schema/output_schema+risk_level+versionBaseAgent（ABC）：declare_capabilities()+execute(params, context)，经CentralHub.register_agent()接入ExecutionResult（dataclass）：success/data/message/suggestions/statusAssemblyState（Pydantic）：全局会话状态（assets/slots/analysis_results/match_results/global_context/...）
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 58 lines · 1,211 tokens per session scan A cd99c2989649
video-clip-agent CLAUDE.md is an instructions file published in the GitHub repository neopen/video-clip-agent (36 stars, last pushed 5d ago), licensed MIT. It adds 1,211 tokens to every session, about $0.0061 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
ai-development-patterns CLAUDE.md
Claude Code instructions for PaulDuvall/ai-development-patterns, covering claude.md, repository overview, repository structure, key architectural principles and pattern architecture.
ai-development-patterns AGENTS.md
AGENTS.md instructions for PaulDuvall/ai-development-patterns, covering agent instructions, task tracking, pattern adoption evaluation and landing the plane (session completion).
awesome-claude AGENTS.md
AGENTS.md instructions for JSONbored/awesome-claude, covering heyclaude agent instructions, project intent, source of truth, contribution and content policy and safety and privacy metadata.
awesome-claude CLAUDE.md
Claude Code instructions for JSONbored/awesome-claude, a project described as: HeyClaude is a curated registry and distribution surface for Claude and AI-workflow assets: agents, MCP servers, skills, commands, hooks, rules, guides, tools, jobs, Raycast feeds, static data exports, and an npm MCP package.
lenserfight AGENTS.md
AGENTS.md instructions for conectlens/lenserfight, covering agents.md, 1. think before coding, 2. simplicity first, 3. surgical changes and 4. performance under stress.
lenserfight CLAUDE.md
Claude Code instructions for conectlens/lenserfight, a project described as: Turn AI prompts into reusable, versioned Lenses - organize with Tags and Threads, automate as Workflows, configure as Agents, evaluate with Battles.