Reads TASKS.md and produces an ordered implementation roadmap with phases, parallel groups, dependency resolution, and checkpoints. Does NOT write code — it generates the plan the developer or AI agent follows.
AI agents call sdd_implement to retrieve information from Specky without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool only reads an existing file (TASKS.md) and generates a planning artifact (roadmap). It explicitly states it does not write code or perform any side-effecting operations, making it a pure read/analysis operation.
From the tool's definition Reads TASKS.md and produces an ordered implementation roadmap... Does NOT write code — it generates the plan the developer or AI agent follows.
Documented attack patterns abuse exactly the kind of access sdd_implement gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Specky, and nothing reaches the server without passing your rules. This is the rule we recommend for sdd_implement:
{
"version": "1",
"default": "deny",
"tools": {
"sdd_implement": {}
}
}sdd_implement 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.
Reads TASKS.md and produces an ordered implementation roadmap with phases, parallel groups, dependency resolution, and checkpoints. Does NOT write code — it generates the plan the developer or AI agent follows. It is categorised as a Read tool in the Specky MCP Server, which means it retrieves data without modifying state.
Register the Specky MCP server in PolicyLayer and add a rule for sdd_implement: 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 Specky. Nothing to install.
sdd_implement 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 sdd_implement 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 sdd_implement. 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.
sdd_implement is provided by the Specky MCP server (paulasilvatech/specky). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Specky, 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.
56 Specky tools catalogued and risk-classified — across an index of 43,000+ MCP servers.