opencode-workbench

Inspect a Linux target

inspect_target

Read-onlyIdempotent

Probe one Linux machine — this host (local) or a remote host over SSH (ssh) — and report its OS, kernel, architecture, package manager, Node/npm, OpenCode, Docker, and per-tool detection flags, plus the names (never values) of credentials present. Read-only: it runs one shell probe as the target user (over SSH when mode:ssh), installs nothing, writes nothing, contacts no external service, and can take a few seconds per hop. Parameter semantics: target.mode selects local vs ssh (default local); target.host is required only for ssh; target.user defaults to the ssh config/current user; target.cwd sets the directory for relative checks; target.port/identityFile pass straight to ssh. Use it to see what a clone would touch, then plan_clone to turn that into an ordered plan and apply_clone to perform it.

Input Schema

Name	Required	Description	Default
`target`	Yes	The machine to operate on: this host (local) or a remote host over SSH (ssh).

Output Schema

Name	Required	Description
`has`	Yes	Detection flags for git, curl, node, npm, corepack, pnpm, docker, uv, gh, opencode, sudo.
`arch`	No	CPU architecture (e.g. x86_64, aarch64).
`home`	Yes	Home directory on the target.
`osId`	No	OS identifier (e.g. ubuntu, debian, rhel).
`user`	Yes	Detected user on the target.
`shell`	No	The target user's login shell.
`kernel`	No	Kernel name and release.
`osName`	No	Human-readable OS name.
`osIdLike`	No	ID_LIKE from /etc/os-release: space-separated distribution family hints.
`configDir`	Yes	OpenCode config directory on the target.
`npmPrefix`	No	npm global prefix, if npm is installed.
`osVersion`	No	OS version string.
`envPresent`	No	Space-separated names of credentials present on the target (values are never read).
`npmModules`	No	npm global modules directory.
`nodeVersion`	No	Installed Node.js version, if any.
`opencodeBin`	No	Path to the opencode binary, if installed.
`skillsCount`	No	Number of SKILL.md files under the target's ~/.config/opencode/skills.
`dockerRunning`	No	Whether the Docker daemon is reachable.
`packageManager`	No	Detected package manager (apt-get, dnf, yum, apk, pacman, zypper, or none).
`opencodeVersion`	No	Version the opencode binary reports, without a leading v (e.g. 2.2.1).
`workspaceDefault`	Yes	Default directory where the profile repo will be cloned.

Schema Changelog

Changes observed during successful MCP inspections.

Changed5 schema fields changedOct 29, 2026

addedOutput schema / properties / npmPrefix

Added value: +{
+  "description": "npm global prefix, if npm is installed.",
+  "type": "string"
+}

addedOutput schema / properties / opencodeVersion

Added value: +{
+  "description": "Version the opencode binary reports, without a leading v (e.g. 2.2.1).",
+  "type": "string"
+}

addedOutput schema / properties / osIdLike

Added value: +{
+  "description": "ID_LIKE from /etc/os-release: space-separated distribution family hints.",
+  "type": "string"
+}

addedOutput schema / properties / shell

Added value: +{
+  "description": "The target user's login shell.",
+  "type": "string"
+}

addedOutput schema / properties / skillsCount

Added value: +{
+  "description": "Number of SKILL.md files under the target's ~/.config/opencode/skills.",
+  "type": "string"
+}

First observedOct 14, 2026

TDQS

A4.8/5.0

Behavior5/5

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

Annotations cover readOnly/idempotent/openWorld, yet the description still adds real operational context beyond them: one shell probe as the target user, installs nothing, writes nothing, contacts no external service, and can take a few seconds per hop. It also discloses a privacy-relevant behavior (credential names only, never values) that no annotation conveys.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the verb and the report contents, then behavior, then parameters, then sibling routing — a sensible order with no filler prose. It is on the long side and the 'Parameter semantics:' recital partially duplicates the schema, costing a little density.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a nested-object, single-parameter probe tool with an output schema, everything an agent needs is present: what is inspected, side-effect profile, latency expectation, SSH vs local behavior, and where to go next. Return-value shape is correctly left to the output schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so baseline is 3, but the description adds meaning the schema does not: mode defaults to local, host is required only for ssh, user falls back to ssh config/current user, cwd governs relative checks, and port/identityFile are raw ssh passthroughs. The only shortfall is that these points largely restate the nested schema rather than adding edge-case behavior (e.g., what happens if host is supplied with mode=local).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Specific verb+resource: 'Probe one Linux machine ... report its OS, kernel, architecture, package manager, Node/npm, OpenCode, Docker, and per-tool detection flags'. It also bounds its own scope (credential names, never values) and positions itself relative to siblings plan_clone/apply_clone, so an agent can separate it from the other inspection-adjacent tools without opening a schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit sequencing guidance: 'Use it to see what a clone would touch, then plan_clone to turn that into an ordered plan and apply_clone to perform it.' It names the downstream alternatives and the condition under which they are used, plus the mode-selection rule (local vs ssh, default local).

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

All opencode-workbench tools Inspect

Try in Browser

Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.