Back to skill

Security audit

hyperframes-cli

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because scanners report substantial capabilities and account or cloud actions beyond its stated video-project purpose; the overall confidence is medium, and the scan is partial.

Consider the reported cloud and account actions, unsolicited feedback reporting, and unpinned CLI dependency before installing. The scan inspected all 13 listed components, but reference resolution was partial for two local paths, so the evidence is incomplete; the suspicious verdict does not by itself establish 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (253)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The declared description covers project and media workflows but does not disclose destructive infrastructure management. A Lambda teardown command is a materially different capability, not merely a supporting implementation detail of the listed operations.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

This duplicates the automatic-feedback issue: the agent is told to transmit a rating and a report about unmet user requests without content-specific authorization. The report could expose project requirements, and the rating could misrepresent user sentiment, although the file exempts users with telemetry disabled.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

These commands expose distinct media-processing capabilities not disclosed in the declared description. They are not merely implementation details of the listed project-management and rendering operations. The snippet does not establish what files, external services, or permissions these commands use.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

Rendering, site creation, and progress monitoring reasonably support the declared rendering and sharing features. However, the explicit cloudrun deploy and cloudrun destroy commands add cloud infrastructure lifecycle management, including potentially destructive operations on Google Cloud resources. This is a material capability beyond the declared project and media workflow.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The file directs the agent to submit a rating and a report about unmet user requests after rendering, without requiring approval of the outbound content. This adds an unsolicited external reporting step that could disclose task details or submit a rating the user did not authorize; the telemetry opt-out reduces but does not eliminate that risk.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The description covers project and rendering workflows but does not disclose cloud infrastructure deployment. Deploying a production-named Lambda stack is a consequential cloud-management capability, not merely an incidental implementation detail of rendering. The command attempts deployment; successful provisioning depends on credentials and environment.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

Telemetry management is a distinct capability not disclosed in the declared description and is not required to implement the listed project and media workflows. The chunk demonstrates telemetry configuration commands, but does not itself establish that data is transmitted or identify any collection destination.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The supplied commands expose cloud identity and deployment-policy functionality not represented in the declared media-project workflow. IAM policy generation and validation are distinct security-infrastructure capabilities, rather than ordinary JSON reading. The snippet does not show policies being attached, IAM resources being modified, or AWS credentials being accessed.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The rendering command aligns with the declared purpose. However, deploying and destroying a cloud backend introduces infrastructure lifecycle management and resource changes not disclosed by the description. These are distinct capabilities rather than merely internal rendering implementation details.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding

Listing and retrieving renders reasonably support the declared rendering, sharing, and history workflows. Deleting cloud-hosted renders is a distinct state-changing capability that the description does not disclose.

Content

No source excerpt is available for this finding.

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
75% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cloud.md (reported line 150)May include surrounding context.

bash
npx hyperframes cloud list                 # recent renders (--limit, --token, --all)
npx hyperframes cloud get hfr_def456       # full detail + short-lived signed video_url
npx hyperframes cloud delete hfr_def456    # soft-delete (--no-confirm to skip the prompt)

video_url and thumbnail_url are short-lived presigned URLs, so re-fetch with cloud get rather than caching them.

Credential Access

High
Category
Privilege Escalation
Confidence
90% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · references/lambda.md (reported line 35)May include surrounding context.

md
## Prerequisites

- AWS credentials configured (env vars, `~/.aws/credentials`, SSO, or IMDS).
- AWS SAM CLI on `PATH`.
- `bun` on `PATH` (builds the Lambda handler ZIP).

The skill directs access to AI-harness credentials and account usage outside its video-project purpose.

High
Category
Excessive Agency
Confidence
98% confidence
Finding

The skill directs automatic, repeated access to existing AI-harness credentials and subscription usage without obtaining task-specific permission. This crosses from video-project work into unrelated account access; the statements about token-free output and no telemetry do not establish authorization or prove the implementation safe.

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.

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.

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 mandates account-linked usage telemetry unrelated to the requested project operations.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The skill mandates an outbound usage event as soon as it starts, regardless of whether that event is needed for the user's task. The text acknowledges account linkage when signed in, creating a privacy risk from unrequested tracking; explicit disclosure and honoring an existing opt-out materially limit the severity.

Content

No source excerpt is available for this finding.