Laup

The actual AI answers for ONE prompt, identified by its fingerprint (the [fp:...] tag used in action evidence). Returns the newest answers with text excerpts and self-mention/citation counts.

get_prompt_responses

CHANGES DATA: move one action card to another triage status (open/planned/doing/measuring/done/wont_do/snoozed). With status=done and verify=true the card is parked in 'measuring' and the next presence run posts a share-of-voice verdict into this thread. Reaching 'measuring' or 'done' also starts outcome tracking: the merchant's presence on the card's prompts over the two weeks before the work is frozen, and every presence run refreshes the change since. Moving a card BACK to open, planned or doing discards that measurement and it cannot be restored: only move a tracked card backwards when the work genuinely needs redoing. This APPLIES immediately: there is no confirmation step and no undo tool. Call it when the person you are working for has asked for this change, not to find out whether they want it. The card records which agent made the change, and a teammate can move a card back or dismiss one you added.

set_action_status

Dated, per-response list of when this domain was cited (grounded on) or self-mentioned by name in an AI answer, newest first -- unlike get_share_of_voice/get_grounding_overview (windowed aggregates), this returns individual events so you can see exactly when a citation or mention happened and on which prompt. Filterable to one engine; pages of 50 via offset/next_offset; n_answers_in_window is the denominator for event rates (events are hit rows only, absences are not listed).

get_citation_events

CHANGES DATA: add a new user-authored action card to the board (lands in the Planned column with the next card number). Use when the merchant wants to track something from this conversation as an action. This APPLIES immediately: there is no confirmation step and no undo tool. Call it when the person you are working for has asked for this change, not to find out whether they want it. The card records which agent made the change, and a teammate can move a card back or dismiss one you added.

create_action_card

How often each AI engine reads web sources before answering, plus the self-credit split: answers that read the merchant's own site but did or did not name the merchant (the uncredited gap). Both aggregate and split per engine (self_credit_per_engine).

get_grounding_overview

Paginated list of the web sources AI engines read/cite when answering this merchant's prompts, with per-source usage counts and whether the source names the merchant. Supports gap filtering (sources used where answers do not name you). Pass 'url' to drill into one source: the result becomes the prompts that caused engines to browse or cite that URL.

get_grounding_sources

One page of the merchant's AI-visibility action plan: recommended and triaged actions with their priority score ('impact', 0-100, higher means do it sooner), status, created/triaged timestamps, who last moved the card (triaged_by), and the reason a 'wont_do' card was declined (wont_do_reason). Lists the whole visible board by default; pass status to see one triage column, and read counts for the per-status totals of the visible board regardless of the filter. A stale 'open' card no longer surfaced by the latest run is excluded from both. Paginated: read n_total, and when has_more is true call again with offset=next_offset to get the rest. Pass include_prompts=true for the prompt fingerprints each action affects (usable with get_prompt_responses); they are omitted by default because they crowd out cards.

list_actions

The team's discussion thread on one action card: who said what, oldest first, including comments an agent posted. Read this before commenting so you do not repeat what someone already said. Each entry carries a comment_id and, for long bodies, a shortened text with a 'truncated' flag. Pass comment_id to re-read one comment in full when its entry came back truncated.

list_comments

List the domains this token can query. The 'domain' field of each row is the exact value to pass as another tool's 'domain' argument, and 'id' works there too. 'country_label' and 'language_label' say which locale that row tracks, which is what separates two rows sharing a hostname. 'status' says whether new measurements are being collected on the regular schedule: 'active' and 'trialing' are; any other value means scheduled collection is stopped or not yet started. It says nothing about the data already collected: every row listed here is fully queryable, and a stopped row keeps its complete history.

list_domains

CHANGES DATA: post a comment into an action card's team discussion thread, under your own agent name. It cannot @mention anyone: notifying teammates stays a human action.

add_comment