# Context Readiness Score

> Readiness is not about how much information the company has. It is whether important context has source, scope, redaction, review, and a route into the next agent session.

## Audit

Score one point for each control that is true today:

- The team can name the calls, tickets, PRs, docs, threads, files, and sessions where agent-relevant context is created.
- Recent product or operating decisions are saved as concise memories, not only buried in source artifacts.
- Every durable memory can point back to the source that produced it.
- The team uses personal, team, project, org, and sensitive scopes deliberately.
- Secrets, credentials, private customer data, and unnecessary personal details are removed before storage.
- Someone reviews stale or conflicting memories before agents rely on them.
- Approved agents receive relevant Ambience context at session start.
- The team has a lightweight review cadence for sweeps, conflicts, scope changes, and promoted skills.

## Interpreting the score

- 0 to 3: cold-start risk. Start with one permissioned source sweep and approve the first durable memories.
- 4 to 6: useful foundation. Seed one project and tighten source, scope, redaction, and review.
- 7 to 8: ready to compound. Turn on regular sweeps and weekly review.

## Why Ambience

Ambience turns this audit into a product workflow: discover available sources, ask what to include, propose memories, show source and scope, redact before storage, and audit agent reads and writes.

## Related

- [Company Context Maturity Model](https://ambience.sh/company-context/maturity-model)
- [Onboarding new agents](https://ambience.sh/company-context/onboarding-new-agents)
- [How to audit AI agent context](https://ambience.sh/answers/how-to-audit-ai-agent-context)