No user-submitted related servers found.
Scored across 42 tools
Several near-duplicate pairs force an agent to guess: meta_get_insights vs meta_get_insights_raw, meta_get_creatives vs meta_get_creative_assets, meta_get_audiences vs meta_get_audience_details, meta_get_ad_studies vs meta_get_study_results vs meta_interpret_experiment_results, and meta_get_node_fields vs meta_list_edge_raw vs meta_get_entity_configuration all overlap heavily. The long descriptions try to carve out niches (curated vs escape hatch), but the boundaries remain blurry at selection time, especially among the raw/generic graph readers.
Almost everything follows a stable meta_verb_noun pattern (get_, list_, search_, validate_, join_, interpret_) with a uniform prefix. A few outliers like meta_health_check, meta_debug_token, and meta_get_uploaded_video deviate slightly but remain readable and predictable.
42 tools is heavy for an ad account surface, and much of the bloat comes from generic escape hatches and one-off field readers (node fields, edge raw, batch status, uploaded video) rather than distinct workflows. The broad domain justifies more than a dozen tools, but not this many near-redundant entries.
Read coverage is genuinely broad across accounts, campaigns, ad sets, ads, creatives, audiences, catalogs, targeting, previews, media library, business assets, and experiments. However, the surface is entirely read-only — no create, update, pause, budget, or delete operations — which is a notable gap if the server is meant to cover the Meta Ads lifecycle rather than serve as a diagnostic/analytics layer.