Diagnostic tools

Use these before a campaign write. They do not pause a campaign, change a budget, or enable spend. They are registered in the catalog and called like any other tool.

Tool	Use it when
`start_here`	You want the safest first steps for this tenant.
`list_what_i_can_do`	You want the capabilities grouped by platform.
`diagnose_my_setup`	You want connection, governance, and recent error readiness for one platform.
`get_campaign_spec`	You need the draft fields and safety rules before filling a campaign in.
`preflight_campaign`	A draft exists and you want it checked before any spend-affecting action.
`validate_campaign_draft`	You want the draft fields and safety defaults validated.
`suggest_next_action`	You want the next read-only step, not a write.
`why_did_this_fail`	A call failed and you have a `trace_id`, `request_id`, or `tool_name`.
`whats_changed_since_last_visit`	You want recent account or service movement since a time you pass in `since`.
`weekly_opportunities`	You want read-only optimization ideas for the week.
`usage_value_summary`	You want usage, governance, and value signals. `get_usage_status` remains the quota row.

why_did_this_fail explains a trace. It does not rerun the write. After you understand the cause, fix the input and approve one corrected attempt. Common causes are listed in Troubleshooting. Platform readiness checks sit next to these helpers: get_google_api_readiness and get_tiktok_api_readiness.

Start

AI clients

Operate

Reference

Ad platforms

Marketing stack

Diagnostic tools