Delete a branch or a ref.
Requires the org's agent-writes switch. Deleting follows the org's review settings for destructive actions (a person's approval by default). The bound repository cannot be deleted.
| Name | Required | Description | Default |
|---|---|---|---|
| name | No | Branch name or git ref | |
| github | No | owner/repo. Omit to use the run's bound github= | |
| target | Yes | branch or ref. The bound repository cannot be deleted |
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Changes observed during successful MCP inspections.
Input schema / properties / name / descriptionPrevious value: -"Branch name, git ref, or omit when target is repo"New value: +"Branch name or git ref"Input schema / properties / target / descriptionPrevious value: -"branch, ref, or repo. repo is the run's bound repository"New value: +"branch or ref. The bound repository cannot be deleted"Input schema / properties / target / enumPrevious value: -[
- "branch",
- "ref",
- "repo"
-]New value: +[
+ "branch",
+ "ref"
+]Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full load and does well: it discloses the write-authorization prerequisite, the approval gate, and a hard guardrail (bound repo undeletable). It stops short of saying whether deletion is reversible or how a missing branch is handled, which is the remaining gap for a destructive mutation.
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?
Four short sentences, front-loaded with the action, then constraints. Every sentence adds distinct information (verb, auth switch, review gating, repo guardrail) with zero repetition.
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?
An output schema exists so return values need no explanation, and the schema fully documents parameters. The description supplies the authorization and safety context an agent needs before calling a destructive tool; only reversibility/error behavior is left unstated.
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?
Schema description coverage is 100% – each of name, github, and target is documented in the schema itself, including the target enum. The description adds no parameter-level detail beyond what structured data already provides, so it sits at the baseline.
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?
States a specific verb and resource ('Delete a branch or a ref') and bounds the scope by declaring the bound repository cannot be deleted. An agent can immediately tell this apart from siblings like github_push, github_merge, or github_close.
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?
Explicitly names the precondition ('Requires the org's agent-writes switch') and warns that deletion is gated by org review settings with human approval by default, which tells the agent when this tool will actually execute. It does not name alternative tools for branch cleanup, but no close sibling offers that.
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.