Reference

You never call these directly — you talk in plain language and Talk2Graph chooses the right action. This page documents what those actions are, the query language underneath, and the settings that govern them, so you understand what's happening and can phrase requests precisely.

Capabilities

Talk2Graph exposes the following capabilities to the assistant. Every one operates on the graph session you have open.

GFQL: the query language underneath

When you ask for something, Louie translates it into GFQL (Graph Frame Query Language) — Graphistry's declarative graph query language — or an equivalent read-only Cypher query. You never have to write it, and when the assistant references the query, it's highlighted as inline code so you can see what ran and refine it. See the full GFQL documentation for the language itself.

A GFQL query is a chain of operations applied left to right:

• Node steps — match nodes by attribute (e.g. status = "active", amount > 1000). An empty match means "all nodes."
• Edge steps — traverse connections, in a direction (forward / reverse / undirected) for one or many hops.
• Call steps — run graph algorithms (PageRank, betweenness, authority score, community detection) or table pipelines (group-by, order-by, limit).

Filters support the usual predicates: comparisons (> < = ≠), ranges (between), set membership (is in), string tests (contains, starts with, ends with, regex match), and null checks. A read-only subset of Cypher (MATCH / WHERE / RETURN / WITH / ORDER BY / LIMIT / CALL) is also accepted — but never write operations (CREATE / MERGE / DELETE / SET).

For example, "color the top influencers" becomes a chain that runs PageRank and colors the top slice:

[
  { "type": "Node" },
  { "type": "Call", "function": "compute_igraph",
    "params": { "alg": "pagerank" } }
]

… and the assistant colors the top ~20% by score.

Limits

The current boundaries, so you know what to expect.