Lesson 4 of 5 · 12 min

Clause risk levels

How MatterFirst flags clauses — categorical levels, not scores.

Lesson 4 of 5

When clause analysis runs, each key clause is tagged with a categorical risk level. There are no numeric risk scores and no configurable thresholds.

Categorical risk levels

Clause analysis asks the model to identify key clauses and label each with one of four levels: info, low, medium or high. These appear as coloured badges next to the clause number, title and one-line summary on the Clauses tab. There is no 0–100 score — the level is the whole story.

No threshold engine

MatterFirst does not let you configure rules like "flag any deposit above 20%" or set per-matter-type alert thresholds — that engine doesn't exist. What you can control is whether clause analysis runs at all: it's a task called "Key Clause Analysis" that you enable in the Analysis Rules tab.

  1. 1

    Enable Key Clause Analysis

    In Settings › Document AI, open the Analysis Rules tab and make sure "Key Clause Analysis" is among the tasks for the relevant document category. This is the switch that produces clause badges; there are no threshold numbers to set.

    Analysis Rules — Key Clause Analysis is a task you toggle, not a threshold engine.
  2. 2

    Read the results in the Clauses tab

    Open an analysed document and switch to the Clauses tab. Each identified clause shows its section reference, a short title, a one-sentence summary, and an info / low / medium / high badge. Skim the higher levels first — they're where the model thinks attention is warranted.

  3. 3

    Interpret the levels

    Treat the levels as triage, not judgement: 'high' means the clause is unusual or consequential enough to read closely, not that anything is wrong. Because there's no configurable threshold, the levels reflect the model's read of the document rather than firm-specific rules.

Tips & pitfalls

  • Risk is categorical — info, low, medium or high. There's no numeric score and no per-matter-type threshold to configure.
  • An empty Clauses tab usually means no notable clauses were found, or Key Clause Analysis isn't enabled for that document category.
  • Clause badges are a reading aid, not legal advice — always open and read anything flagged medium or high yourself.

Key takeaways

  • Clause risk is a categorical label — info, low, medium or high.
  • There is no numeric risk score and no configurable threshold engine.
  • Enable 'Key Clause Analysis' in the Analysis Rules tab to get clause badges.
Custom extraction rules Connecting AI to workflows

Put the lesson into practice.

Book a guided session and our team will help you configure MatterFirst for your own practice.

Book a session Back to the Academy