Deploy SAM
All Features Project Memory

One set of project rules for every agent

SAM tells new sessions to load your rules and preferences first, and you can edit what's saved.

  • Project memory entries carry a confidence score, a source, and a confirmation date
  • Policies are injected into every agent session as instruction-only context
  • A full activity log of every task, session, and execution event
Read the guide
Agent Context overview showing 25 memory entities, 12 active policies, and 50 recent actions, with the context stack agents receive: repository instructions, memory, policies, and profilesAgent Context overview showing 25 memory entities, 12 active policies, and 50 recent actions, with the context stack agents receive: repository instructions, memory, policies, and profiles

Persistent project memory

Store observations about your architecture, conventions, and preferences. Each entry has a confidence score, a source (explicit or inferred), and a confirmation date so you can see what the knowledge base contains.

Memory tab listing knowledge entities such as Architecture and Code Quality, each with observations showing a confidence percentage, an explicit or inferred source, and a confirmation dateMemory tab listing knowledge entities such as Architecture and Code Quality, each with observations showing a confidence percentage, an explicit or inferred source, and a confirmation date Each observation shows confidence, source, and when it was last confirmed

Project policies

Define project-level instructions like "use portals for modals" or "CLI packages must meet QA standards." Policies are tagged as preferences or rules and injected into every agent session as context.

Policies tab listing project policies such as Never push to main and Refund changes require a Stripe replay test, each tagged rule, constraint, or delegation with a confidence and marked instruction-onlyPolicies tab listing project policies such as Never push to main and Refund changes require a Stripe replay test, each tagged rule, constraint, or delegation with a confidence and marked instruction-only Policies are instruction-only — injected into agent context, not programmatically enforced

Activity log

A timestamped feed of every task submission, execution step, session start/stop, and agent completion event. Filter to find specific activity across the project.

Agent actions tab showing a timestamped feed of task.agent_completed, task.execution_step, session.started, comment.created, and session.stopped events with their session topicsAgent actions tab showing a timestamped feed of task.agent_completed, task.execution_step, session.started, comment.created, and session.stopped events with their session topics Every event type — task completions, execution steps, session lifecycle changes
Previous Full Visibility & Control Next Idea to Execution

Ready to try SAM?

SAM is open source. To deploy it, you need a Cloudflare Workers Paid plan, a domain on Cloudflare, and a GitHub account.

Deploy SAM Read the Docs