get_session_summary
What did session X actually do? AI-generated 2–4 sentence summaries.
How to use it
get_session_summary is exposed by the Observer MCP server. Add the server to your MCP client (Claude Desktop, Cursor, Windsurf and others), and the get_session_summary tool becomes available to the model automatically. See the full listing for setup details and every tool this server provides.
Install Observer
npx @superbased/observerOther tools in Observer (20)
Live prompt-cache health: which caches are warm, expiring, or cold, with value-at-risk.
Did this exact command already run? With what result?
Has this file changed since I last read it?
A distilled, scrubbed handover of a session from another AI tool, so you can continue its work here.
The raw row, scrubbed of secrets.
Per-window spend rollup.
Error correlation + retry detection.
Every read/edit of this file across every tool + session (with codeintel enrichment when available).
Without re-running.
Evidence-backed model suggestion per turn-kind, from the local Model Value Report.
Code vs. explanation split of a session's output, by bytes, with the code:explanation ratio and languages used.
Derived behaviours: hot files, co-changes, edit→test pairs.
What would Discovery flag for this project?
Model-routing layer state: phase, available policy templates, tier-table size, decision-log counters.
One full, un-excerpted message from a session's transcript — pulls the complete body a handover excerpt truncated.
For resuming an interrupted session.
Top dollar/time-quantified cost & quality suggestions from the local advisor.
Chronological ±N actions around an actionid.
Full-text search of past tool-call outputs (FTS5 over excerpts).
Fuzzy symbol search across the project's codeintel index (Tier-C).