Skip to content
Optra Prism Docs

Search is only available in production builds. Try building and previewing the site to test it out locally.

PRISM Score v3.0

PRISM Score v3.0 is a from-scratch rebuild of the personal score. Instead of blending five inputs into a Skill composite, v3.0 asks one direct question for every session:

Did this session crush its goal?

Your PRISM Score is the percentage of qualifying sessions that did.

PRISM Score = 100 × crushed_count / total_count

A session “crushes” when four facts all land:

  1. Substance floor passed — the session did real work (≥3 turns, or ≥10 net LOC, or ≥1 mutating tool call)
  2. Goal complete — verified by an LLM outcome judge against per-intent criteria
  3. Not rework — no later session reverts or rewrites the same code
  4. A clear intent was established — the rubric judge committed to a class

Sessions that fail the substance floor or never establish a goal are excluded from both numerator and denominator. Trivial chat doesn’t drag your score down, and it doesn’t pad it either.

Why a new score

Section titled “Why a new score”

v2.1’s Skill composite (SSE · PES · IE · CRR · FC) measured prompt behavior. v3.0 measures outcomes. Two developers can have identical prompt rubrics and very different real-world results — v3.0 catches that.

The full rationale, side-by-side metrics, and migration notes live on the comparison page.

Session labels

Section titled “Session labels”

Every session also gets two human-readable labels:

  • Title — up to 60 characters (e.g. “Fix off-by-one in JWT parser”).
  • Summary — a 1–2 sentence recap, up to 240 characters.

Labels follow a three-step lifecycle:

  1. Heuristic — when the session opens, a title is taken from a truncation of the first user prompt.
  2. Rubric — when the session closes, the rubric LLM call rewrites both fields with a cleaner title and summary.
  3. User-edited — you can rename a session in the dashboard at any time. A user edit locks the session against future overwrites.

What’s in this section

Section titled “What’s in this section”

Status

Section titled “Status”

v3.0 ships behind a dashboard toggle (PRISM Score v3). Both scores run side-by-side during the calibration period. Once v3.0 reaches steady-state agreement against the hand-labeled set, v2.1 retires.