langchain-reelspan

LangChain and LangGraph tools for Reelspan. Let an agent publish videos, photos and text to TikTok, YouTube, Instagram, Facebook, X, Threads, LinkedIn, Pinterest, Bluesky, Reddit and more, schedule them, check each network's result and read analytics.

pip install langchain-reelspan
export REELSPAN_API_KEY=rs_live_...   # dashboard → Developers → API keys

from langchain.agents import create_agent
from langchain_reelspan import ReelspanToolkit

tools = ReelspanToolkit(profile="acme").get_tools()
agent = create_agent("anthropic:claude-sonnet-5-5", tools)

agent.invoke({"messages": [{"role": "user", "content":
    "Post https://cdn.example.com/launch.mp4 to TikTok and YouTube tomorrow 9am Berlin time with a short caption."}]})

Tools

Publishing tools take scheduled_date + timezone, add_to_queue, first_comment and settings for network options ({"privacy_level": "PUBLIC_TO_EVERYONE", "reddit_subreddit": "videos"}; Upload-Post field names work too).

Options

from reelspan import Reelspan

ReelspanToolkit(
    client=Reelspan(api_key="rs_live_..."),  # default: reads REELSPAN_API_KEY
    profile="acme",                # pin to one profile; the model can't post as another
    platforms=["linkedin", "x"],   # only these networks
    read_only=True,                # only tools that read
    wait=True,                     # wait for each network's result after posting (default)
)

Single tools work too: ReelspanPublishText(profile="acme").invoke({"platforms": ["x"], "caption": "Hi"}).

Errors (a wrong key, a network that isn't allowed, a missing profile) come back to the model as tool messages it can act on. Posts carry an idempotency key, so a retried call never double-posts.

Links

API reference · Python SDK · MCP server

MIT licensed. Security reports: see SECURITY.md.