agentmods.dev

vibe-video CLAUDE.md

Project instructions for Vibe Video, a conversational team of software agents that creates Hyperframes motion-graphics videos from plain-language prompts. They describe the project's architecture, agents, tools, and source of truth.

In plain words
What is it for?
Use them when modifying the Vibe Video team, its animator, code explorer, or researcher agents, the API server, video tools, settings, database, or project specification.
Why use it?
They give coding agents the context needed to work consistently in an unfamiliar video-generation codebase and keep its specification aligned with the implementation.

Instructions file

102 4mo ago A scan 1,501 tokens original Apache-2.0
View the source file agno-agi/vibe-video
Install

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.

agentmods
npx agentmods add instructions/agno-agi/vibe-video/claude-md
Clone the repo
git clone --depth 1 https://github.com/agno-agi/vibe-video
Per session 1,501 This file is loaded in full into every session.
When invoked 1,501 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.01501 $0.01501
Opus 5 $0.00750 $0.00750
Sonnet 5 $0.00300 $0.00300
Haiku 4.5 $0.00150 $0.00150

Measured yesterday against content hash 3ee8977300cd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

vibe-video 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 yesterday.

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.

CLAUDE.md · 139 lines

How it starts

The opening of the file, as written. The whole thing — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Vibe Video

A conversational multi-agent team that produces Hyperframes motion-graphics videos from natural-language prompts. Part of the "mini demo agent" family alongside vibe-to-prd.

The canonical specification is docs/SPEC.md. All other documentation derives from it — keep the two in sync.

Architecture

  • Team definition: vibe_video/team.py (Vibe Video team leader, Coordinate mode)
  • Member agents: vibe_video/agents/animator.py (Animator), vibe_video/agents/code_explorer.py (CodeExplorer), vibe_video/agents/researcher.py (Researcher, optional)
  • Shared settings: vibe_video/settings.py (DB, REPOS_DIR, RENDERS_DIR, model)
  • API server: app/main.py (FastAPI + AgentOS web UI)
  • Custom tools: vibe_video/tools/git.py (GitTools), vibe_video/tools/hyperframes.py (HyperframesTools)
  • Hyperframes authoring guide: compact, hand-written, inlined directly in vibe_video/agents/animator.py (no external skill assets)
  • Database: PostgreSQL + pgvector (session memory + agentic memory)

Team Structure

Vibe Video (Team, Coordinate, claude-opus-4-7)
├── Animator   — writes Hyperframes HTML, renders to MP4, iterates
├── CodeExplorer   — clones arbitrary repos on demand, read-only
├── Researcher — web search via Parallel (optional; gated on PARALLEL_API_KEY)
└── [leader responds directly for greetings/simple questions]
  • Leader — routes, delegates, synthesizes. No tools of its own.
  • Animator — HyperframesTools, FileTools, ReasoningTools.
  • CodeExplorer — CodingTools (read-only), GitTools (clone + read), ReasoningTools.
  • Researcher — ParallelTools, ReasoningTools. Instantiated only when PARALLEL_API_KEY is set.

Key Concepts

  • Coordinate mode: leader picks the right specialist, delegates a structured brief, synthesizes. Respond directly only for greetings and "what can you do?".
  • Hyperframes: HTML + GSAP local-render framework from HeyGen. The Animator writes index.html + optional compositions/*.html into a per-session workspace under RENDERS_DIR, lints, renders to MP4 via the CLI, and iterates on failure (up to MAX_RENDER_RETRIES).
  • On-demand cloning: CodeExplorer has no pre-configured repo list. REPOS_DIR is an ephemeral scratch dir; clone_repo(url) is idempotent (clones or fast-forwards).
  • Skill-in-prompt: the Hyperframes and GSAP skills from heygen-com/hyperframes are fetched at build time and inlined into Animator's system prompt.
  • Agentic memory: captures user style preferences (e.g. "3Blue1Brown pacing", narration on/off) across sessions. No learnings KB.
  • Scheduler: scheduler=True in AgentOS. No pre-registered jobs — the feature is surfaced so users can schedule their own recurring runs against the team from the web UI.

Read the full file on GitHub · 139 lines

Changes

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.

  1. yesterday First seen · 139 lines · 1,501 tokens per session scan A 3ee8977300cd

Subscribe to this mod's changes

vibe-video CLAUDE.md is an instructions file published in the GitHub repository agno-agi/vibe-video (102 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 1,501 tokens to every session, about $0.0075 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.

Backend and APIsDatabasesResearchagnoai-agentsvideo
Related

Other instructions, from other repositories

mediago AGENTS.md

Instructions for mediago-dev/mediago, covering repository guidelines, project structure & module organization, tooling and script placement, documentation site safety and build, test, and development commands.

mediago-dev/mediago · 1,037 tokens

mediago copilot-instructions.md

Instructions for mediago-dev/mediago, covering github copilot instruction: code review and optimization, role and objective, review perspective and principles and specific instructions.

mediago-dev/mediago · 653 tokens

openstory CLAUDE.md

Instructions for openstory-so/openstory, covering claude.md, commands, dev, quality and tests.

openstory-so/openstory · 9,211 tokens

dash CLAUDE.md

Instructions for agno-agi/dash, covering claude.md, project overview, structure, commands and data & knowledge.

agno-agi/dash · 2,286 tokens

yutu AGENTS.md

Instructions for eat-pray-ai/yutu, covering yutu, quick reference, directory index, documentation and conventions.

eat-pray-ai/yutu · 564 tokens

scout AGENTS.md

Instructions for agno-agi/scout, covering agents.md, project overview, architecture, contextprovider and one write surface.

agno-agi/scout · 3,932 tokens