Connect

Memories

Confirmed findings that agents and people save so the next investigation starts from what you already know, plus the notes Polylane writes each day.

Memories are confirmed findings about your system, saved by agents and by you so the next investigation starts from what you already know. Notes are Polylane's own daily summaries.

What it does

A memory is a markdown document with a title and labels that records one finding: a confirmed failure mode, a recurring pattern, or an operational fact backed by evidence. Agents search memories when they investigate, so a root cause found once is not rediscovered. Agents do not store user preferences, speculative conclusions, or configuration already visible in the topology.

How memories are created

When an investigation confirms a root cause or a recurring pattern, the agent saves it with saveMemory, recording the finding, the evidence, the affected resources and the thread it came from. Scripts can do the same through POST /v1/memories.

To write one yourself, open Settings > Memories and select New memory, then add markdown content, labels and an optional title. From a thread, ask in plain words: "Save a memory: checkout-api integration tests fail intermittently because the payment vendor's sandbox times out under parallel load, confirmed from pipeline logs showing 504 responses."

How memories reach a thread

Polylane indexes every memory for search by meaning and by keyword. Agents search workspace memory whenever they investigate an issue and pull in matching passages. You can also attach a memory to a thread, and its content goes straight into the conversation.

Saved once, that finding lets the next agent investigating a checkout-api test failure rule out the known failure mode before digging further. Write memories the same way: name the service, the failure mode, the evidence and the affected resources. Use Edit, Archive or Delete on a memory's page when reality changes.

Notes

Once a day Polylane summarizes the day's thread activity into a daily note, and updates a single workspace note only when the day reveals a fact that will still be true in a month. Notes come from threads shared with the workspace or public; days with no thread activity produce no note. Every new thread's briefing includes the workspace note and the three most recent daily notes, and you can edit, copy, download or delete both kinds.

KindWritten byDescriptionReach for it when
MemoryAgents or people, one finding at a timeA specific, evidence-backed fact with labels and provenanceA finding should survive across investigations
Workspace notePolylane, daily, editableDurable facts about the workspace: architecture, tooling, conventions, ownershipYou want to see what agents are told about your workspace
Daily notePolylane, one per day, editableDay-level patterns and notable signals, not a play-by-playYou want a record of what a day's activity looked like
  • Threads for where memories are searched and saved.
  • Fix runs for the investigations that produce confirmed findings.
  • Topology for what agents read from the graph instead of memory.