Memory & Context

Build a compressed symbol map of a TypeScript or Swift repo for LLM context

Install a memory classifier that keeps preferences across agent sessions

Summarize recent work and project state to restore context after clearing a chat

Write a structured handoff doc with files, decisions and next steps for another agent

Load a handoff document and resume work where the last session stopped

Update an existing handoff document with progress, status and next steps

Wrap verbose git, test and CLI output with RTK to save context tokens

Save decisions, changed files and next steps to a handoff file for resuming

Measure fixed context token overhead in your agent config and get fixes

Keep plans, progress and notes in files on disk for long multi-step tasks

Search and analyze past session logs with jq to recover earlier context

Extract reusable patterns from coding sessions and save them as learned skills

Compact context at logical task breaks instead of letting auto-compaction cut

Hand the current conversation to a fresh background agent that picks up the work

Condense the current conversation into a handoff doc for another agent to pick up

Look up earlier compacted messages and sub-agents after context is lost

Compress a conversation into a structured context snapshot on request

Run an Obsidian vault as a self-updating second brain with scheduled maintenance

Install, search and maintain the Agent Memory Vault shared memory system

Configure lossless context management for OpenClaw with summary health and recall

Capture key decisions and a session summary to Open Brain when a session ends

Automatically save session transcripts to Open Brain with session-end hooks

Mirror local agent memory files into Open Brain whenever a memory is saved

Guard Open Brain thought deletions, since hard deletes cannot be undone

Keep agent memory traceable with provenance, scope, review and use policies

Capture and search Open Brain thoughts over plain HTTPS when MCP is blocked

Correct, retag or supersede existing Open Brain thoughts without creating new ones

Write and maintain daily memory files and MEMORY.md for agent workspaces

Turn documents into a queryable knowledge graph of entities and relations

Run a blameless root-cause analysis after a mistake and record a durable lesson

Add and search memories as a tool throughout a Codex session

Keep persistent memory across sessions in a Basic Memory knowledge graph

Save and recall Pi session checkpoints in Basic Memory for continuity

Save a Codex handoff checkpoint in Basic Memory with an exact resume command

Log engineering decisions with rationale, alternatives and consequences

Resume Codex work from a Basic Memory checkpoint or current repo evidence

Save a quick fact, reminder or preference to Basic Memory from Codex

Share a personal Basic Memory note to a team project with explicit confirmation

Apply a customizable writing standard to Basic Memory notes that Codex creates

Create, run and check agents that automate work on Basic Memory notes

Capture a thread's state into one Basic Memory note that updates in place

Write Basic Memory notes with live charts, tables and boards using Comark components

Resume prior work by rebuilding context from the Basic Memory graph

Link orphan notes, merge duplicates and build hub notes in a knowledge graph

Split bloated notes, resolve contradictions and retire stale content in memory

Turn transcripts and pasted documents into structured notes with action items

Archive finished notes and move them between status folders with frontmatter updates

Build a structured knowledge graph from a full literary work, chapter by chapter

Search notes by frontmatter fields like status, priority or confidence