runtime_peers_digest
Read runtime_state_digest from two or more peers over the SAME root and field set and report whether they agree (read-only). Converged means every peer's digest is byte-equal; when they differ, diverged_at names the node paths that disagree. The sequence that actually converges, in this order: ru...
This record as markdown: /tools/breakpoint-mcp/runtime-peers-digest.md
What runtime_peers_digest does on Breakpoint
AI agents call runtime_peers_digest to retrieve information from Breakpoint without modifying anything. It is typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
root | string | Yes | Root node path to digest in each peer (same path on every peer) |
peers | array | — | Peer ids to compare (default: every live peer; at least two are required) |
fields | array | — | Property names to capture per node (default: the runtime_state_digest defaults) |
max_depth | integer | — | Max recursion depth (default 8) |
Parameters from the server's own tool schema.
Why runtime_peers_digest is rated Low
The tool is explicitly described as read-only, comparing digests across peers to check convergence without modifying any state. It retrieves and compares data, making it a Read category tool with low severity.
From the tool's definition Read runtime_state_digest from two or more peers over the SAME root and field set and report whether they agree (read-only)
Risk signalsBulk/mass operation — affects multiple targets · Admin/system-level operation
Attacks that exploit this kind of access
The rule that runs runtime_peers_digest safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Breakpoint, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For runtime_peers_digest, this is the rule to start with:
runtime_peers_digest is read-only, so it stays allowed. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Breakpoint, apply this rule, and every runtime_peers_digest call is checked against it from then on.
Questions about runtime_peers_digest
Read runtime_state_digest from two or more peers over the SAME root and field set and report whether they agree (read-only). Converged means every peer's digest is byte-equal; when they differ, diverged_at names the node paths that disagree. The sequence that actually converges, in this order: runtime_spawn_peers, then per peer runtime_time_scale{scale:0} to FREEZE FIRST, runtime_set_property{peer} to equalise the starting state, runtime_seed_rng{seed} with the same seed, runtime_step_frames{frames:K,kind:"physics"}, then this. FOUR PRECONDITIONS, every one measured on real Godot 4.3 (convergence re-measured on 4.7) — a run that skips any of them diverges, and (2) is the one that surprises people. (1) Step the FIXED physics timestep: state advanced on the variable idle-frame delta is real elapsed wall-clock time in each process and never converges, so pass kind:"physics". (2) The global RNG must be consumed ONLY on frames you are stepping. runtime_seed_rng seeds one stream shared by the whole project, and freezing does NOT stop it being drawn — time_scale 0 zeroes delta but callbacks still fire, so unconditional draws burn the stream at wall-clock rate while frozen, and idle-frame draws burn it during the step. Guard draws on delta > 0 and give idle-frame code its own RandomNumberGenerator; otherwise even fixed-timestep physics state diverges. (3) Peers free-run between spawn and freeze for different durations, so their state already differs before you begin: freeze first, then equalise with runtime_set_property{peer}. (4) Same machine only — peers share one OS and one engine build, and this claims nothing about convergence across machines. It is categorised as a Read tool in the Breakpoint MCP Server, which means it retrieves data without modifying state.
runtime_peers_digest accepts 4 parameters: root, peers, fields, max_depth. Required: root. The full parameter table on this page comes from the server's own tool schema.
Register the Breakpoint MCP server in PolicyLayer and add a rule for runtime_peers_digest: 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 Breakpoint. Nothing to install.
runtime_peers_digest 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 runtime_peers_digest 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 runtime_peers_digest. 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.
runtime_peers_digest is provided by the Breakpoint MCP server (breakpoint-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Breakpoint, and thousands of servers like it.
This server
Across the catalogue