16 docs tagged with "AI Agents"

The three kinds of agents in FunnelStory — Outcome Agents, built-in agents, and agents you build — and how each detects, investigates, and acts.

AI Agents are configurable automations in FunnelStory that run a series of steps: pull data from your workspace or connections, call an LLM where you need judgment or language, then take action (Slack, email, tasks, CRM updates, datasets, and more). You choose when they run—on a schedule, when events occur, when SQL returns matching rows, or only when someone starts a run manually or from chat.

AI summaries turn long threads, tickets, and usage context into short, sourced explanations attached to each needle mover. They answer “why is this here?” without asking you to open every underlying document first.

Agents FunnelStory builds and tunes for one job: investigating needle movers and drafting Action Cards, and preparing you for customer meetings.

Custom Datasets are named, workspace-scoped tables that agents can read from and write to across runs. They are the persistent memory layer for AI Agents — the place where a run stores what it computed, checked, or decided so the next run can pick up where the last one left off.

These patterns are starting points—swap table names, thresholds, Slack channels, and connection IDs for your workspace. Each example lists the intent, trigger, and shape of the graph; paste fragments into the canvas or provide them to an assistant following Vibe coding.

Create FunnelStory agent definitions (flows) — JSON configurations you edit in the agent builder or export/import that define multi-step data processing and LLM agent workflows.

CALL steps invoke functions by function_id. Each function validates its arguments; failures surface as step errors (the run stops along that path).

This walkthrough takes you from an empty workspace to a saved, testable agent that reads account data, summarizes it with an LLM, and sends a Slack message—so you learn triggers, CALL steps, and an AGENT step in one pass.

An AGENT step runs an LLM with a system prompt, a user prompt, optional tools, and optional multi-turn behavior. Use it whenever you need summarization, classification, routing decisions, or natural-language output—while keeping deterministic work in CALL steps.

Each step in an agent graph has an op that tells the runner what to do next. Steps share common fields: id (must match the key in the steps map), next (empty string "" ends the path), optional out for where results are stored, and an op-specific block (call, agent, loop, etc.).

The Step through flow panel runs an agent one step at a time, like a code debugger. Instead of firing the whole graph and reading logs afterward, you pause before every step, inspect the exact variables that step will see, edit them if you need to, and then advance.

FunnelStory gives you step-level tests, trigger previews, Chat tryouts, and a run history so you can ship agents confidently.

Triggers define when FunnelStory starts a new run for an agent. A published (non-draft) agent with a configured trigger runs automatically; draft agents and agents with no trigger only run when someone starts them manually.

Steps pass data through variables. Templates turn those values into SQL, JSON arguments, and prompts.

Vibe coding means describing the agent you want in everyday language and letting an AI assistant (Cursor, Claude Desktop, etc.) draft the JSON, call APIs, and run tests on your behalf. FunnelStory supports this through the MCP server and a published flow authoring guide your assistant can read.