Back to skill

Security audit

Appian Missingdescr

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small, purpose-aligned Appian audit wrapper, but users should validate the application UUID and quote command arguments before running it.

Install only if you already trust the APPIAN_RUNNER script and Appian environment you will point it at. Before running, validate that the application identifier is truly a UUID and prefer a quoted or shell-free invocation such as passing the runner path and UUID as fixed arguments rather than interpolating raw text into a shell command.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:25
Finding

Unquoted User-Controlled Input in Shell Command

Content
View full analysis

Vulnerability Details

File Location: SKILL.md:25-29
Vulnerability Type: Command injection through unquoted and unvalidated shell arguments
Risk Level: High

bash
node $APPIAN_RUNNER missing-descr APPLICATION_UUID

Replace APPLICATION_UUID with the UUID the user provided.

Technical Analysis

The documented command incorporates the user-provided application UUID without requiring validation or shell-safe argument handling. The APPIAN_RUNNER environment variable is also expanded without quotation or path validation.

If an agent replaces APPLICATION_UUID through textual command construction and invokes the resulting string using a shell, shell metacharacters can terminate or alter the intended command. For example, a supplied value containing ;, command substitution, redirection, or other shell syntax could cause unintended commands to execute. A malicious or compromised APPIAN_RUNNER environment value could similarly inject additional shell arguments or commands.

The project does not include the external runner implementation, so its internal behavior could not be assessed. Exploitability is conditional on the agent executing the documented command through a shell rather than through a shell-free argument-array API.

Attack Path

  1. An attacker provides a purported application UUID containing shell syntax, such as valid-looking-value; attacker-command.
  2. The agent follows the instruction to replace APPLICATION_UUID with the supplied value.
  3. The agent constructs the documented command as a shell command without validation or quoting.
  4. The shell interprets the metacharacters rather than treating the entire UUID as a single literal argument.
  5. The appended command executes with the same operating-system identity, filesystem access, environment access, and network permissions as the agent process.

A second path exists if an attacker can influence APPIAN_RUNNER: the unquoted va ...[truncated 655 chars]

Remediation
View remediation

Remediation Suggestions

  1. Validate the supplied identifier against the exact Appian application UUID syntax before execution. Reject every value that does not fully match the expected format and length.
  2. Invoke Node.js through a shell-free process API with a fixed argument array, for example conceptually:
    text
    executable: node
    arguments: [validatedRunnerPath, "missing-descr", validatedApplicationUuid]
    shell: false
    
  3. If a shell is unavoidable, quote both substitutions:
    bash
    node "$APPIAN_RUNNER" missing-descr "$APPLICATION_UUID"
    
    Quoting should supplement, not replace, strict UUID validation.
  4. Resolve APPIAN_RUNNER to a canonical path and require it to match an approved local runner file. Reject paths outside the trusted installation directory.
  5. Ensure the runner and its parent directories are not writable by untrusted users.
  6. Update the skill instructions to prohibit direct textual interpolation of user input into shell commands.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep