Alerting Rules Engine

What it does

Generates deterministic alerts from latest vs prior result comparison.

Why this is insightful

Teams stop reacting late. Alerts surface meaningful rank drops/disappearances quickly with evidence.

Required API fields

• query.keyword
• rankings.brands[].name
• rankings.brands[].rank
• createdAt

Optional input

• alertConfig

How to use

1. Compare current and prior rank states.
2. Evaluate threshold rules.
3. Assign severity and include field-level evidence.

Expected output

• alert list with severity
• message
• evidence payload

Copy-ready skill prompt

Use this as a full copy/paste prompt in your AI tool:

Skill: Alerting Rules Engine
Goal: Run this analysis on FullMention data only.
Data policy:
- Do one controlled ingestion pull from FullMention API.
- Store dataset locally (file/local DB) and analyze offline.
- No repeated API fetch loops during analysis.
Execution contract:
- Provide ingestion_plan, max_api_calls, cache_key, dataset_timestamp, analysis_mode=offline_only.
Required fields:
- Use the required API fields listed on this page.
Output requirements:
- Return summary, key metrics, and evidence_map with source_field_paths and sample_result_ids.
- Include limitations and confidence score rationale.