list_audit_events
List what has been done on this organization’s behalf, newest first.
The audit timeline: mutations made from the dashboard, changes the agent
made in chat, suggestions that were accepted and applied, integrations
connected or disconnected — plus changes detected in the ad platform’s own
console, marked with an integration actor so they are distinguishable
from anything done here.
Top-level actions only. When one authorized action fanned out into several
platform writes (an accepted suggestion pausing three ad sets), the parent
row reports the count and the children are reachable by parent_id.
Args: since: Optional ISO-8601 datetime; only actions at or after this time actor: ‘user’ | ‘agent’ | ‘system’ | ‘integration’ action: Action key filter surface: Where the action came from search: Free-text match against the rendered summary limit: Max events (default 50, max 200) offset: Pagination offset
Returns: A dictionary containing:
- events: [{id, occurred_at, summary, action, status, error, actor, surface, target, ad_account, platform, changes, parent_id}]
- total_count, limit, offset, has_more
- source: provenance tag
Parameters#
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
since | string | no | Optional ISO 8601 datetime; only actions at/after it. | |
actor | string | no | Filter by who acted: ‘user’, ‘agent’, ‘system’, ‘integration’. | |
action | string | no | Filter by action key, e.g. ‘ads.entity.status_changed’. | |
surface | string | no | Where it came from: ‘dashboard’, ‘chat’, ‘suggestion’, ‘mcp’, ‘slack’, ‘system’. | |
search | string | no | Free-text match against the rendered summary. | |
limit | integer | no | 50 | Max events to return (default 50, max 200). |
offset | integer | no | 0 | Pagination offset (default 0). |
include_details | boolean | no | false |
Generated by generate_mcp_docs from the server tool. Add to this page below this block; the block is rewritten on every run.