set_milestonesReplace the list of ordered milestones under the goal (2–6, or up to 6 per track), each countable (metric and target) or measured by its steps, with a due date before the goal deadline; due dates in order within a track. Pass each existing id, or that milestone's steps are deleted. To change one milestone, use edit_milestone.
| Name | Required | Description | Default |
|---|---|---|---|
| milestones | Yes |
Changes observed during successful MCP inspections.
Input schema / properties / milestones / items / properties / trackIdAdded value: +{
+ "description": "the track this milestone belongs to (set_tracks); required for open milestones once the path has tracks",
+ "type": "string"
+}Input schema / properties / milestones / items / properties / kindAdded value: +{
+ "description": "count: progress is a recorded value against target (\"talk to 20 people\"); steps: progress is steps done (\"implement the onboarding flow\"), no metric needed. Default: count when metric and target are given, else steps; an existing milestone keeps its kind",
+ "enum": [
+ "count",
+ "steps"
+ ],
+ "type": "string"
+}Input schema / properties / milestones / items / requiredPrevious value: -[
- "title",
- "metric",
- "target",
- "due"
-]New value: +[
+ "title",
+ "due"
+]Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already flag destructiveHint=true and idempotentHint=true, but the description supplies the consequence that matters: omitting an existing id deletes that milestone's steps. That is the key non-obvious destructive detail an agent cannot infer from the annotations alone.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Everything present earns its place and the replacement semantics are front-loaded, but it is a single dense run-on sentence built from semicolons and parentheticals, which is harder to parse than it needs to be for a constraint-heavy tool.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a single-array-parameter, destructive replace operation with no output schema, the description covers limits, ordering, id semantics, and the alternative tool. The remaining minor gap is not stating explicitly that milestones omitted from the array are removed, beyond the implied consequences of 'Replace' and the id rule.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description adds real meaning beyond the schema: the 2–6 count limit, the six-per-track cap, the due-before-goal-deadline and within-track ordering rules, and the 'pass each existing id' contract that governs create-vs-update vs delete. It leaves metric/target values largely to the schema's own nested descriptions, so it is strong but not exhaustive.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Opens with a specific verb and resource ('Replace the list of ordered milestones under the goal') and scopes the operation as a full replacement rather than a per-item edit. It explicitly names the sibling edit_milestone as the tool for single-milestone changes, so an agent can route without opening either schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
States the selection condition against an alternative directly: 'To change one milestone, use edit_milestone.' It also encodes the applicable constraints (2–6, or up to 6 per track, due before the goal deadline, due dates ordered within a track), so the agent knows both when this tool applies and what input will be rejected.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.