agenticow_branch
agenticow@~0.2.3 — COW-fork a base .rvf memory file. Measured 162-byte branches regardless of base size (verified at N=1k/10k/50k). Use when you need per-Darwin-iteration / per-user / per-session memory personalization. Copying the parent .rvf file is wrong because full-copy snapshots grow linear...
This record as markdown: /tools/io-github-ruvnet-claude-flow/agenticow-branch.md
What agenticow_branch does on Claude Flow
AI agents use agenticow_branch to create or update resources in Claude Flow, usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Claude Flow environment.
Why agenticow_branch is rated Medium
This tool creates a new copy-on-write branch (fork) of a memory file, which is a Write operation — it creates a new data artifact (a 162-byte branch file) derived from a parent. It does not execute code, delete data, or involve finances. The severity is medium because misuse could proliferate forked memory states across sessions/users, but the operation is reversible and scoped to memory personalization files.
From the tool's definition COW-fork a base .rvf memory file... Use when you need per-Darwin-iteration / per-user / per-session memory personalization
Attacks that exploit this kind of access
The rule that runs agenticow_branch safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Claude Flow, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For agenticow_branch, this is the rule to start with:
agenticow_branch stays usable, but capped: an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Claude Flow, apply this rule, and every agenticow_branch call is checked against it from then on.
Questions about agenticow_branch
agenticow@~0.2.3 — COW-fork a base .rvf memory file. Measured 162-byte branches regardless of base size (verified at N=1k/10k/50k). Use when you need per-Darwin-iteration / per-user / per-session memory personalization. Copying the parent .rvf file is wrong because full-copy snapshots grow linearly (the 3.3 GB Darwin-worktree bloat fixed in v3.14.4); agenticow gives read-through semantics (parent ∪ edits, child wins) at constant 162 B. Optional dep — degrades to {degraded:true} when missing. It is categorised as a Write tool in the Claude Flow MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Claude Flow MCP server in PolicyLayer and add a rule for agenticow_branch: 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 Claude Flow. Nothing to install.
agenticow_branch is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the agenticow_branch 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 agenticow_branch. 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.
agenticow_branch is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Claude Flow, and thousands of servers like it.
This server
Across the catalogue