frameworks

A set of small connectors that expose Watch Skill's video-watching, question-answering, and search functions to several AI agent frameworks. These frameworks are libraries for building software agents.

In plain words
What is it for?
Use it with LangChain, LangGraph, CrewAI, OpenAI Agents SDK, LlamaIndex, or AutoGen to summarize videos, ask questions, and search indexed recordings.
Why use it?
It avoids writing a separate video integration for each framework while keeping the same underlying functions.

Agent

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 agents/oxbshw/watch-skill/frameworks
Clone the repo
git clone --depth 1 https://github.com/oxbshw/watch-skill
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,203 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. 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 $0.00000 $0.01203
Opus 5 $0.00000 $0.00602
Sonnet 5 $0.00000 $0.00241
Haiku 4.5 $0.00000 $0.00120

Measured 2d ago against content hash cf61128e05ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

frameworks scanned grade B with 1 finding 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 2d 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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

const res = await fetch("http://127.0.0.1:8748/v1/ask", { method: "POST",
docs/agents/frameworks.md · 135 lines

How it starts

The opening of the file, as written. The whole thing — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Framework adapters

MCP already covers Claude Code/Desktop, Cursor, Cline, Windsurf, Codex, Gemini CLI, and VS Code (see the per-agent guides in this directory). For agent frameworks, Watch Skill ships thin native adapters — three tools (watch_video, ask_video, search_videos) wrapping the same core calls. Adapters never add logic; everything stays in the engine.

Install the engine plus the extra for your framework:

pip install "watch-skill[langchain]"     # or crewai / openai-agents / llamaindex / autogen

LangChain / LangGraph

from watch_skill.integrations.langchain import get_watch_tools

agent = create_agent(model, tools=get_watch_tools())
agent.invoke({"messages": [("user", "Watch https://youtu.be/… and summarize it")]})

Tools are langchain_core StructuredTools — anything consuming langchain-core tools (LangGraph included) uses them unchanged. Runnable proof: examples/09-framework-adapters/langchain_example.py.

CrewAI

from watch_skill.integrations.crewai import get_watch_tools

analyst = Agent(role="video analyst", goal="answer questions about videos",
                backstory="…", tools=get_watch_tools())

Runnable proof: examples/09-framework-adapters/crewai_example.py.

OpenAI Agents SDK

from agents import Agent
from watch_skill.integrations.openai_agents import get_watch_tools

agent = Agent(name="video analyst", tools=get_watch_tools())

Runnable proof: examples/09-framework-adapters/openai_agents_example.py.

LlamaIndex

from watch_skill.integrations.llamaindex import get_watch_tools

agent = FunctionAgent(tools=get_watch_tools(), llm=llm)

Tools are llama_index.core.tools.FunctionTool objects (verified against the LlamaIndex tools documentation; adapter unit-tested).

AutoGen (v0.4+)

from autogen_agentchat.agents import AssistantAgent
from watch_skill.integrations.autogen import get_watch_tools

agent = AssistantAgent(name="analyst", model_client=client, tools=get_watch_tools())

Read the full file on GitHub · 135 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. 2d ago First seen · 135 lines · 0 tokens per session scan B cf61128e05ac

Subscribe to this mod's changes

frameworks is an agent published in the GitHub repository oxbshw/watch-skill (320 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,203 tokens. A static security scan graded it B with 1 finding (sends data to an external url). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.