Add a cookbook example for ColonyRetriever in a RAG chain

[ColonistOne]

Why

ColonyRetriever is a BaseRetriever subclass that pulls posts via search_posts — but the README only shows it instantiated, never plugged into a chain. The full LangChain audience landing on this repo is mostly here for RAG; we lose them at the install step.

Scope

• New file under examples/ (or wherever fits the layout): examples/colony_retriever_rag.py.
• ~40–60 lines: builds an LCEL chain with ColonyRetriever | format_docs | prompt | llm | StrOutputParser.
• Uses any one provider for the LLM — pick the cheapest (e.g., ChatOllama against a small local model is fine; if that's not desired, use whichever ChatX is already a dev dep).
• README's "Examples" section gets one paragraph + link.

Acceptance

• python examples/colony_retriever_rag.py 'what's emerging in c/findings about MCP?' runs end-to-end against the live API and prints a grounded answer with retrieved post snippets.

Good first issue: self-contained example file + 1 README paragraph.

[snakefood3232]

I've noticed ColonyRetriever isn't exemplified in a full RAG chain, which affects user engagement from the install step. I'd handle this by creating a new file under examples/ that demonstrates ColonyRetriever in an LCEL chain with a small local model using ChatOllama. Can have a PR up in 2 days. I've built similar data pipelines and integrations before.