Back to skill

Security audit

PaleBlueDot CLI

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly coherent documentation for a CLI, but it recommends an unsafe remote installer and under-warns users about sensitive session cookies and API token keys.

Review before installing. Avoid running the one-line installer unless you trust the upstream repository and have independently inspected or verified the script. Prefer a pinned release or checksum-verified install. Treat session cookies and token keys as passwords, do not paste them into shared chats or logs, and only delete or reveal token keys after explicitly confirming the target token ID.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:21
Finding

Unverified Remote Installation Script Executed Directly by Bash

Content
View full analysis
Remediation
View remediation
pbd-cli.tar.gz" | sha256sum -c - ``` 5. If a shell installer remains necessary, download it to a local file, verify its signature or checksum, and allow inspection before execution. 6. Prefer a trusted package manager or a signed platform-specific package where feasible. 7. Avoid requiring administrative privileges. Install into a user-controlled directory such as `~/.local/bin` by default, and clearly explain any optional system-wide installation step. 8. Include the installer or its security-relevant source in the audited Skill package so reviewers can assess its behavior. 9. Use `curl -f` so HTTP failures are not silently passed into the shell, though this does not replace integrity verification. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

Piping curl output directly into bash creates a command chain that removes any opportunity to inspect or validate the fetched content before execution. In an agent environment, this is especially dangerous because it normalizes immediate execution of remote content and can turn a documentation step into arbitrary code execution on the host.

Content

Scanner excerpt · SKILL.md (reported line 21)May include surrounding context.

Linux / macOS:

bash
curl -sSL https://raw.githubusercontent.com/PaleBlueDot-AI-Open/pbd-cli/main/install.sh | bash

This script automatically detects your OS and architecture, downloads the latest release, and installs to /usr/local/bin.

External Model or Provider Selection

High
Category
Excessive Agency
Confidence
90% confidence
Finding

Skill selects an external model or provider that may use a different account or billing plan than the operator expects. Undisclosed model switches can cause unexpected cost or quota consumption.

Content

Scanner excerpt · SKILL.md (reported line 208)May include surrounding context.

bash
pbd-cli usage logs
pbd-cli usage logs --limit 50 --model gpt-4o
pbd-cli usage logs --page 2 --token my-token

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manual login flow instructs the user to provide a session cookie and user ID but does not warn that the session cookie is an active authentication credential equivalent to account access. In a skill used by an agent, this omission materially raises the risk of prompting users to paste sensitive credentials into chat, terminals, or logs where they may be exposed.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill documents destructive token deletion with no warning, confirmation, or advice to verify the token ID before execution. In an agent-assisted context, this increases the chance of accidental credential revocation or disruption of production integrations if the command is run on the user's behalf.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill exposes a command that retrieves a token's plaintext key without warning that the output is highly sensitive. In an agent setting, this can lead to accidental display, logging, transcript capture, or onward disclosure of secrets that can be used to access the user's account or API resources.

Content

No source excerpt is available for this finding.

External Script Fetching

Low
Category
Supply Chain
Confidence
97% confidence
Finding

The one-line installer downloads a remote script from GitHub and immediately executes it via the shell, with no integrity verification, pinning, or review step. If the upstream repository, network path, or referenced script is compromised, users could execute arbitrary code with their local privileges.

Content

Scanner excerpt · SKILL.md (reported line 21)May include surrounding context.

Linux / macOS:

bash
curl -sSL https://raw.githubusercontent.com/PaleBlueDot-AI-Open/pbd-cli/main/install.sh | bash

This script automatically detects your OS and architecture, downloads the latest release, and installs to /usr/local/bin.