Skip to main content
Glama

HALLUX

by dev.blvkware
OverviewDiscussionsAdmin

Check an install command before running it

hallux_check_command
Read-onlyIdempotent

Call this BEFORE running any command that installs a dependency: npm, yarn, pnpm, bun, pip, uv, poetry, pipx, cargo, go get, dotnet add package, nuget.

Language models invent plausible package names that have never existed, and attackers register those names because they can predict them. Installing one runs attacker code. You cannot tell the difference by looking at the name, which is why this check exists.

Pass the command exactly as you intend to run it. The identifiers are extracted for you.

If the result says BLOCK, do not run the command. Use the successor if one is given, otherwise tell the user what was found and stop.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
commandYesThe shell command, verbatim. Example: pip install requests requests-oauth2-helper

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
blockedNoThe absent, phantom and squat results.
checkedNoHow many identifiers were checked.
commandNoThe command as received.
resultsYesEvery result, in input order.
decisionYesBLOCK: do not proceed. REVIEW: something could not be verified. PROCEED: every identifier exists. NOTHING_TO_CHECK: no identifier was recognised.
unverifiedNoThe unknown results.
safeToProceedYesFalse whenever anything was blocked.
identifiersFoundNoThe identifiers read off the command.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
+  "properties": {
+    "blocked": {
+      "description": "The absent, phantom and squat results.",
+      "items": {
+        "properties": {
+          "confidence": {
+            "maximum": 1,
+            "minimum": 0,
+            "type": "number"
+          },
+          "evidence": {
+            "description": "What the verdict rests on: registry status and time, attestations, model families, nearest existing names.",
+            "type": "object"
+          },
+          "namespace": {
+            "description": "The namespace checked.",
+            "type": "string"
+          },
+          "reason": {
+            "description": "Why the verdict is unknown.",
+            "type": "string"
+          },
+          "recommendation": {
+            "description": "What to do next.",
+            "type": "string"
+          },
+          "successor": {
+            "description": "The name to use instead, when one is known.",
+            "type": "string"
+          },
+          "value": {
+            "description": "The identifier checked.",
+            "type": "string"
+          },
+          "verdict": {
+            "description": "exists and deprecated may proceed; absent, phantom and squat must not; unknown is never permission.",
+            "enum": [
+              "exists",
+              "deprecated",
+              "absent",
+              "phantom",
+              "squat",
+              "unknown"
+            ],
+            "type": "string"
+          }
+        },
+        "required": [
+          "namespace",
+          "value",
+          "verdict",
+          "confidence"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "checked": {
+      "description": "How many identifiers were checked.",
+      "minimum": 0,
+      "type": "integer"
+    },
+    "command": {
+      "description": "The command as received.",
+      "type": "string"
+    },
+    "decision": {
+      "description": "BLOCK: do not proceed. REVIEW: something could not be verified. PROCEED: every identifier exists. NOTHING_TO_CHECK: no identifier was recognised.",
+      "enum": [
+        "PROCEED",
+        "REVIEW",
+        "BLOCK",
+        "NOTHING_TO_CHECK"
+      ],
+      "type": "string"
+    },
+    "identifiersFound": {
+      "description": "The identifiers read off the command.",
+      "items": {
+        "properties": {
+          "namespace": {
+            "type": "string"
+          },
+          "value": {
+            "type": "string"
+          }
+        },
+        "required": [
+          "namespace",
+          "value"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "results": {
+      "description": "Every result, in input order.",
+      "items": {
+        "properties": {
+          "confidence": {
+            "maximum": 1,
+            "minimum": 0,
+            "type": "number"
+          },
+          "evidence": {
+            "description": "What the verdict rests on: registry status and time, attestations, model families, nearest existing names.",
+            "type": "object"
+          },
+          "namespace": {
+            "description": "The namespace checked.",
+            "type": "string"
+          },
+          "reason": {
+            "description": "Why the verdict is unknown.",
+            "type": "string"
+          },
+          "recommendation": {
+            "description": "What to do next.",
+            "type": "string"
+          },
+          "successor": {
+            "description": "The name to use instead, when one is known.",
+            "type": "string"
+          },
+          "value": {
+            "description": "The identifier checked.",
+            "type": "string"
+          },
+          "verdict": {
+            "description": "exists and deprecated may proceed; absent, phantom and squat must not; unknown is never permission.",
+            "enum": [
+              "exists",
+              "deprecated",
+              "absent",
+              "phantom",
+              "squat",
+              "unknown"
+            ],
+            "type": "string"
+          }
+        },
+        "required": [
+          "namespace",
+          "value",
+          "verdict",
+          "confidence"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "safeToProceed": {
+      "description": "False whenever anything was blocked.",
+      "type": "boolean"
+    },
+    "unverified": {
+      "description": "The unknown results.",
+      "items": {
+        "properties": {
+          "confidence": {
+            "maximum": 1,
+            "minimum": 0,
+            "type": "number"
+          },
+          "evidence": {
+            "description": "What the verdict rests on: registry status and time, attestations, model families, nearest existing names.",
+            "type": "object"
+          },
+          "namespace": {
+            "description": "The namespace checked.",
+            "type": "string"
+          },
+          "reason": {
+            "description": "Why the verdict is unknown.",
+            "type": "string"
+          },
+          "recommendation": {
+            "description": "What to do next.",
+            "type": "string"
+          },
+          "successor": {
+            "description": "The name to use instead, when one is known.",
+            "type": "string"
+          },
+          "value": {
+            "description": "The identifier checked.",
+            "type": "string"
+          },
+          "verdict": {
+            "description": "exists and deprecated may proceed; absent, phantom and squat must not; unknown is never permission.",
+            "enum": [
+              "exists",
+              "deprecated",
+              "absent",
+              "phantom",
+              "squat",
+              "unknown"
+            ],
+            "type": "string"
+          }
+        },
+        "required": [
+          "namespace",
+          "value",
+          "verdict",
+          "confidence"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    }
+  },
+  "required": [
+    "decision",
+    "safeToProceed",
+    "results"
+  ],
+  "type": "object"
+}
  2. First observed

TDQS

A4.6/5.0
Behavior4/5

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

Annotations already declare readOnlyHint=true and idempotentHint=true, but the description adds critical behavioral context: the tool performs extraction of identifiers, outputs a result that can be BLOCK, and provides a 'successor' field. It also explains the security rationale, which goes beyond the annotations. The description does not contradict the annotations, and the extra details justify a high score, though it could mention what happens when the result is not BLOCK (e.g., it's safe to proceed) which is why it's not a 5.

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?

The description is well-structured with a clear call-to-action first, followed by rationale, usage instructions, and handling of results. It is a bit longer than necessary, but every paragraph serves a purpose. The security rationale is important context but could be condensed. Overall, it's effective and front-loaded.

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?

Given the tool's simplicity (single parameter, no nested objects) and the presence of an output schema, the description is complete. It covers the purpose, usage timing, alternative actions, and safety handling. The agent has everything needed to invoke it correctly without additional information.

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?

The schema fully describes the 'command' parameter with an example, so the baseline is 3. The description adds crucial semantic context: it specifies that the command should be passed exactly as intended, that identifiers are extracted automatically, and that the tool is for install commands. This enhances understanding beyond the schema, justifying a 4.

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?

The description clearly states the verb ('check') and the resource ('install command'), and explicitly lists the package managers it applies to. It differentiates from siblings like hallux_check and hallux_check_manifest by focusing specifically on install commands, and provides a concrete example, leaving no ambiguity about what the tool does.

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?

The description explicitly states when to use this tool: 'BEFORE running any command that installs a dependency' and lists all relevant package managers. It also gives clear guidance on what to do with the result (e.g., if BLOCK, do not run; use successor if given). This is exemplary usage guidance.

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

All HALLUX toolsInspect
Try in Browser
Shareable, User-Owned Context BackendPeople, applications, and agents read, write, and coordinate together in one context lake. Reach it over MCP, a CLI, a Python SDK, or REST.fulcradynamics.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources