Compare current repository state against the last scan snapshot. Detects changed, added, or deleted files - especially package.json, Dockerfile, CI config, and .env.example. Returns which CLAUDE.md sections may need updating.
AI agents call check_drift to retrieve information from RepoMemory without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
check_drift is a diagnostic tool that queries and compares repository state with a baseline snapshot. It identifies drift (changes/additions/deletions) and reports which sections need updates, but takes no action itself. This is purely a read operation with no side effects, fitting the Read category with low severity since misuse would only expose stale analysis rather than cause harm.
From the tool's definition Tool description emphasizes "Compare current repository state against the last scan snapshot" and "Returns which CLAUDE.md sections may need updating." The verbs are passive/informational: detect, compare, return.
Attacks that exploit this kind of access
Compare current repository state against the last scan snapshot. Detects changed, added, or deleted files - especially package.json, Dockerfile, CI config, and .env.example. Returns which CLAUDE.md sections may need updating. It is categorised as a Read tool in the RepoMemory MCP Server, which means it retrieves data without modifying state.
Register the RepoMemory MCP server in PolicyLayer and add a rule for check_drift: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches RepoMemory. Nothing to install.
check_drift is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the check_drift rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for check_drift. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
check_drift is provided by the RepoMemory MCP server (praveenjayaprakash-jp/repomemory). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →