Advisor & Session Summaries

Prism generates two distinct LLM-authored artifacts to help you understand and improve a session. Both live in the dashboard — there is no in-terminal advisor in the plugin.

Session summaries

After a session’s telemetry lands, the Prism worker produces a short narrative of what happened: the goal you started with, the path the turns took, the outcome, and the two or three turns that mattered most.

Summaries appear on the session detail page and inside the coaching cards on your report. They aren’t a transcript — they’re meant to be scannable in 10 seconds.

The advisor

The Advisor produces per-sub-session coaching cards — short, specific suggestions keyed to the prompt-rubric checks that failed most often on that sub-session. Each card has:

• The weakness (e.g. “Concrete references — no file path or symbol was given”)
• One concrete habit change (“add file paths and a line number”)
• A rewrite of a prompt from the session that would have scored higher

Advisor output is generated by the Prism worker and shows up in the dashboard:

• The Task Advisor drawer on the Realtime view — per sub-session, with the weak point and an example rewrite, as sessions are scored.
• On your Report under “Worst prompts” and the habits chapter.

There is no plugin/in-terminal advisor and no inline coaching printed in your Claude Code session — the coaching is read on the dashboard after a sub-session is scored.

Scoring is independent

Session summaries and advisor cards are generated on top of scoring — they don’t drive it. The prompt rubric runs independently, so the absence of a summary or advisor card for a session doesn’t change its grade.