Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- check_drip_structure — Check the SHAPE of a drip flow for defects decidable from the definition alone: no exit at all, an exit set that omits the flow's own goal so converters keep receiving it, two steps scheduled at the same hour, a step past a hard time exit that can never send, a trigger that is also an exit, and steps listed out of delay order. Findings split by whether the flow can MISDELIVER or is merely oddly shaped. No severity score, no benchmark, no earnings estimate.
- read_cart_recovery — Report abandoned-cart recovery by segment, and attribute an effect to your email ONLY where you supplied a holdout. Without one the recovery rate is reported and nothing is credited to the campaign - some proportion of carts return on their own and land in the same numerator. Coverage is reported only against an emailable count you state, and is never ranked as an opportunity: low coverage is usually a consent fact, not money on the table. A rate needs at least ${MIN_CARTS_FOR_RATE} carts behind it.
- read_subscription_transitions — Count what happened when subscriptions changed plan, and report the MEAN MRR step already observed for each upgrade trigger with its standard error beside it. Nothing is forecast: there is no estimated uplift, and triggers are ordered by COUNT rather than by value, because ranking by money tells the reader that promoting the trigger would earn it. A trigger with fewer than ${MIN_EVENTS_FOR_MEAN} events gets a count and no mean. A cancellation with no stated reason is excluded from the shares, never bucketed.
- run_full — Run every capability you supplied the input for, in registry order, and return one report. Capabilities whose input is missing are listed as skipped with the fields they need — never run against a fixture and never reported as zero.
- list_capabilities — List what this agent can do: every tool it advertises, what each one is for, and the arguments each one reads. Useful for discovery before the first call.
- plan_inputs — Plan the inputs for a tool: returns the questions to ask, the JSON schema of the arguments, and a ready-to-edit example. Every argument it names is an argument the tool actually reads.
- discover_intent — Understand your goal and co-design the exact input through clarifying questions before anything runs.
- open_form — Use this the moment the user wants to run this agent on their own input. It opens an interactive form that collects the run arguments. Do NOT check credentials, do NOT ask clarifying questions in chat, and do NOT run any analysis before calling this — the form is the entry point for running the agent.
- run_form — Internal: invoked by the open_form form when the user submits. Reads exactly the same arguments as run_full and forwards all of them. Not for direct use.
