Back to skill

Security audit

slideshow

Security checks for vulnerabilities and agentic risk

Overview

The scanners rate this slideshow skill suspicious because it requires unrelated telemetry and has supply chain and third-party code risks; the review is incomplete because a referenced file was only partially inspected.

Consider whether you want to allow the required telemetry, unpinned package commands, and runtime code from a public CDN. The scan is incomplete: only one of two components was fully inspected, and static analysis of the reference was limited. The suspicious verdict signals review-worthy risks, not proof of malicious intent.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (11)

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
85% confidence
Finding

Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Content

Scanner excerpt · SKILL.md (reported line 60)May include surrounding context.

md
## Present and hand off

- **Never render a deck to MP4**: `render` keeps only the first scene and truncates silently. For stills, `snapshot --at` each slide.
- Run `npx hyperframes check --json` until it is clean. The slideshow rule checks that every `sceneId` resolves, every hotspot `target` exists, fragments sit inside their slide, and main-line slides don't overlap.
- Serve it with `npx hyperframes present <project-dir>` and hand over the URL it prints: "Click Present, or press P, to open the audience tab."
- The project's root `index.html` should open as a working deck: `<hyperframes-slideshow>` around `<hyperframes-player src="…">`, with the raw composition in a subfolder if needed. The component owns Present (P), Prev/Next, the counter, global mute, fullscreen, swipe, hotspots, presenter notes (editable, kept per deck) and stopping media on slide exit. Don't build your own versions of any of these.
- Screen sharing: in Google Meet share the audience **tab**, which keeps renderin

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 64)May include surrounding context.

md
- A deck opened from the bare player bundle without `present` needs workarounds: `references/standalone-harness.md`.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

The skill requires a startup telemetry beacon unrelated to creating or presenting slides.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

The skill mandates an outbound usage event before performing the requested slideshow work, potentially associating skill usage with a signed-in account. Disclosure and honoring an opt-out reduce the risk, but the instruction still causes unnecessary collection unless the user has already consented to this telemetry.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx hyperframes'.

Medium
Category
MCP Rug Pull
Confidence
83% confidence
Finding

Reviewed independently, this reference supplies a bare npx hyperframes command without a version or a local-wrapper resolution rule. If the package is not available locally, executing it as written can download and execute a changing registry package with the agent's privileges; it is a CLI supply-chain risk, not an MCP-specific issue.

Content

No source excerpt is available for this finding.

The Three.js setup loads and executes third-party CDN code without disclosing the privacy and trust implications.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The generated composition imports executable JavaScript from a public CDN at runtime, making deck execution depend on that provider and exposing request metadata to it. The exact Three.js version reduces version drift but does not verify the delivered bytes; compromised delivery could run code with the composition's origin privileges, including access to same-origin resources.

Content

No source excerpt is available for this finding.

Catalog searches are required to use English without user opt-in or a documented justification.

Low
Category
Output Handling
Confidence
94% confidence
Finding

The prescribed catalog command explicitly requires describing the visual "in English." This imposes a specific language on generated search text without offering a choice or explaining why the catalog requires English.

Content

No source excerpt is available for this finding.