Run a command here and receive a digest of its output

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

All four annotation hints are false, so the description carries the full behavioral burden — and it pays it in full: execution environment (bash -c, this user's privileges, stdin closed), return shape (digest plus a trailer with turn id, artifact ref, exit code, raw size), raw output persisted as a sliceable artifact, and server-side secret expansion where placeholders 'never appear in the result.' Nothing contradicts the annotations.

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 sentences with no filler, and the core action plus execution context are front-loaded before usage routing and the secret-handling caveat. It is on the longer side, but for a six-parameter tool with digest, artifact, and secret mechanics, each sentence carries distinct decision-relevant information, so the density is earned.

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?

This is a high-complexity tool (arbitrary shell execution with side-effect and security implications), and the description covers execution environment, alternative routing, return/artifact flow, and secret handling; an output schema exists to document return values. Minor gaps remain — no explicit statement about non-zero exit handling or how it relates to local_llm_delegate — though the trailer's exit-code field implies graceful handling.

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 coverage is 100% with unusually strong per-parameter descriptions (verbatim even explains when not to use it), establishing a baseline of 3. The description adds one schema-absent semantic: placeholders such as [SECRET-1] in the command are expanded server-side and stripped from results, which meaningfully extends the command parameter's value space beyond the schema's examples.

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?

The description opens with a precise action — 'Execute a shell command on this host (bash -c, this user's privileges, stdin closed) and have the local worker model digest the output for you' — naming the verb, resource, and the differentiating mechanism (digestion by the local worker model). It also explicitly references the sibling local_llm_artifact for the stored raw output, so an agent can distinguish this tool from the sibling set without opening any 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?

The description gives explicit routing: 'Use this instead of your own shell tool whenever the command would print more than you need to read', backed by a concrete rule of thumb ('anything over ~40 lines / 2 KB of output belongs here, while commands that print little or nothing run directly') and examples (logs, tests, builds, listings, git output, grep results). The verbatim parameter schema description adds a further exclusion ('NOT for questions, counts, summaries or listings'), making when-to-use versus when-not-to-use unambiguous.

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