Read or control an owned local job

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare destructiveHint=true, idempotentHint=false and openWorldHint=true, so the safety profile is partly covered; the description adds valuable non-annotation context: output is untrusted data and must never be treated as instructions, plus cursor semantics (UTF-16 offsets for stdout/stderr). It stops short of saying what 'stop' terminates irreversibly or that operations differ in side effects, which matters for a mixed read/destroy tool.

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 dense sentences with no filler; the operation surface and the untrusted-output warning are front-loaded. Packing cursor/UTF-16 and file-index details mid-paragraph is efficient, though the single block is heavier than it needs to be for scanning.

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?

A ten-operation, seven-parameter tool with no output schema and no annotations on per-operation behavior needs more than this; the description omits how to form 'points' for breakpoints, what 'mode' selects during stepping, and what status/diagnostics actually return. The injection warning and cursor guidance cover the highest-risk areas, but the calling details remain incomplete.

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 0% across 7 parameters, so the description carries the full burden. It explains cursor and hints at file indices ('File indices follow runtime source path order'), but job, target, sourceRevision, mode (into/over/out) and points (the breakpoint coordinate array) are never explained anywhere, leaving the most error-prone parameters undocumented.

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+resource with the full operation surface enumerated (status, paged stdout/stderr/diagnostics/locals, stop, step, continue, breakpoints) and scopes it to 'this grant's matching job'. An agent can distinguish this from siblings like crabrix_runtime_request or crabrix_runtime_start without opening the 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?

Gives a real prerequisite (Step/Continue/breakpoints 'require its paused Debug session') and explicit exclusions ('No arbitrary watch expressions or shell commands'). It does not name which sibling to use when the job is not owned by this grant (e.g., crabrix_runtime_clients or crabrix_runtime_request), so the routing is clear but not complete.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.