AI agents call context_pack to retrieve information from Code Graph Knowledge System without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Based on the tool's name and the overall system design as a queryable knowledge graph, 'context_pack' most likely packages or retrieves contextual information from the Neo4j graph for code analysis. Sibling tools show the server has clear Read (query-based), Write (add_*), and Execute (ingest, impact analysis) operations.
From the tool's definition Tool name 'context_pack' combined with sibling tools that include 'add_*', 'batch_extract_*', 'code_graph_impact', and 'code_graph_related'. The naming suggests aggregation or packaging of context/query results from the knowledge graph.
Documented attack patterns abuse exactly the kind of access context_pack gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Code Graph Knowledge System, and nothing reaches the server without passing your rules. This is the rule we recommend for context_pack:
{
"version": "1",
"default": "deny",
"tools": {
"context_pack": {}
}
}context_pack is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
context_pack. It is categorised as a Read tool in the Code Graph Knowledge System MCP Server, which means it retrieves data without modifying state.
Register the Code Graph Knowledge System MCP server in PolicyLayer and add a rule for context_pack: 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 Code Graph Knowledge System. Nothing to install.
context_pack 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 context_pack 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 context_pack. 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.
context_pack is provided by the Code Graph Knowledge System MCP server (royisme/codebase-rag). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Code Graph Knowledge System, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
30 Code Graph Knowledge System tools catalogued and risk-classified — across an index of 43,000+ MCP servers.