Security audit

Lightpanda Browser

Security checks for vulnerabilities and agentic risk

Overview

This skill is clearly about Lightpanda browser usage, but its install instructions ask users to run mutable nightly binaries and Docker images without integrity verification.

Review before installing. Prefer a pinned Lightpanda release or Docker image digest with publisher-provided checksum or signature verification. Avoid running the nightly binary or container in sensitive directories or with mounted credentials until you trust the upstream source and version.

A.I.G

Vulnerability Patterns

Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Insecure DependenciesIntroduces malicious components through unsafe dependency sources
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

Findings (2)

T03 · Remote Payload Retrieval and Execution

Error

Location: SKILL.md:24
Finding: Mutable External Binary Is Downloaded and Executed Without Integrity Verification
Content: View full analysis
Remediation: View remediation
" EXPECTED_SHA256="" curl --fail --location --proto '=https' \ -o lightpanda \ "https://github.com/lightpanda-io/browser/releases/download/${LIGHTPANDA_VERSION}/lightpanda-x86_64-linux" printf '%s %s\n' "$EXPECTED_SHA256" lightpanda | sha256sum --check --strict - chmod 0755 ./lightpanda ``` The expected digest must come from a trusted, immutable publisher source and must not be fetched from the same mutable location without independent authentication. ]]>

T08 · Insecure Dependencies

Warning

Location: SKILL.md:98
Finding: Docker Deployment Uses an Unpinned Nightly Image
Content: View full analysis
Remediation: View remediation
``` 2. Obtain the digest from authenticated publisher metadata and verify image signatures where supported. 3. Document a controlled process for reviewing and updating the pinned digest. 4. Run the container as a non-root user if the image supports it. 5. Drop unnecessary capabilities with `--cap-drop=ALL`, then add back only those demonstrably required. 6. Use `--security-opt=no-new-privileges`, a read-only root filesystem, and resource limits where compatible. 7. Avoid mounting the Docker socket, host credentials, or sensitive directories into the container. 8. Restrict outbound connectivity if browser operation does not require unrestricted network access. ]]>

SkillSpector

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

Findings (2)

Rp1

Medium

Category: MCP Rug Pull
Confidence: 75% confidence
Finding: Docker image references without a specific tag (:latest is implicit) or digest (@sha256:...) can be silently replaced by a malicious image.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The manifest description and heading mix English with Chinese and present the primary body text in Chinese, but the file does not explicitly offer a language/locale choice or state that the skill is region-specific. This can violate language/locale policy when users are not given an opt-in for the imposed language preference.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.