Back to skill

Security audit

Holiday Homework Voice Pack

Security checks for vulnerabilities and agentic risk

Overview

This skill is not clearly malicious, but it needs Review because it asks for broad Beatra account powers and can silently replace its own package files after installation.

Install only if you are comfortable giving this package a broad shared Beatra device authorization and allowing silent package updates by default. Before using it, consider disabling auto-updates with the documented command and verify that the Beatra account scopes, billing controls, and permitted tools match what you want for a homework voice generator.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • 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)

T05 · Unauthorized Access and Privilege Escalation

Error
Location
scripts/authorize.py:34
Finding

Overprivileged Device Authorization Combined with Unrestricted MCP Tool Dispatch

Content
View full analysis
dict[str, Any]: session = _session_with_registration( state_dir=Path.home() / ".beatra", post_json=_default_post_json, ) if command == "tools": return session.request(2, "tools/list", {}) try: arguments = json.load(os.sys.stdin) except json.JSONDecodeError as exc: raise RuntimeError("Tool arguments on stdin must be one JSON object") from exc if not isinstance(arguments, dict): raise RuntimeError("Tool arguments on stdin must be one JSON object") assert tool_name is not None return session.request( 2, "tools/call", {"name": tool_name, "arguments": arguments}, ) ``` ### Technical Analysis The declared functionality is a holiday-homework speech generator. It legitimately needs text-to-speech, selected voice operations, limited artifact upload, task-status access, and potentially billing information. The authorization request nevertheless includes unrelated image, video, and music generation capabilities, along with wallet spending and task cancellation. This violates least privilege because compromise or misuse of the shared bearer credential exposes capabilities that are not necessary to produce homework audio. The risk is amplified by `_run_command`, which accepts an arbitrary MCP tool name and forwards it to `tools/call` without a package-level allowlist. The remote service may still enforce the token's granted scopes and its own tool schemas. How ...[truncated 1647 chars]
Remediation
View remediation

T03 · Remote Payload Retrieval and Execution

Warning
Location
scripts/mcp_client.py:969
Finding

Silent Retrieval and Replacement of Executable Skill Files Without Independent Signature Verification

Content
View full analysis
dict[str, Any]: discovery = _json_object( get_bytes( _discovery_url(), UPDATE_DISCOVERY_TIMEOUT_SECONDS, MAX_UPDATE_DISCOVERY_BYTES, ), "Beatra update discovery", ) current = _semver(PACKAGE_VERSION) available = _semver(discovery.get("version")) _release_urls(discovery) if available < current: raise RuntimeError("Beatra update discovery attempted a version downgrade") return { "current_version": PACKAGE_VERSION, "available_version": discovery["version"], "update_available": available > current, "discovery": discovery, } ``` The remotely supplied hashes are then used to validate the downloaded manifest and archive: ```python def download_update( discovery: dict[str, Any], *, get_bytes: GetBytes = _default_get_bytes, ) -> tuple[dict[str, Any], dict[str, bytes]]: archive_url, manifest_url = _release_urls(discovery) manifest_content = get_bytes( manifest_url, UPDATE_DOWNLOAD_TIMEOUT_SECONDS, MAX_UPDATE_MANIFEST_BYTES, ) if _sha256(manifest_content) != discovery["manifest_sha256"]: raise RuntimeError("Beatra update manifest checksum does not match discovery") manifest = _json_object(manifest_content, "Beat ...[truncated 5503 chars]
Remediation
View remediation
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (39)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill documentation references uninstall and credential/state revocation behaviors that affect shared local files and other installed skills, even though the manifest describes audio generation. Blending lifecycle management and shared state manipulation into a content-production skill creates a risk of unintended deletion, credential revocation, or disruption of unrelated tooling.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill documentation references uninstall and credential/state revocation behaviors that affect shared local files and other installed skills, even though the manifest describes audio generation. Blending lifecycle management and shared state manipulation into a content-production skill creates a risk of unintended deletion, credential revocation, or disruption of unrelated tooling.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill documentation references uninstall and credential/state revocation behaviors that affect shared local files and other installed skills, even though the manifest describes audio generation. Blending lifecycle management and shared state manipulation into a content-production skill creates a risk of unintended deletion, credential revocation, or disruption of unrelated tooling.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill describes silent automatic self-update, download, verification, replacement, rollback, and recovery behavior unrelated to homework audio generation. Any mechanism that can replace package-owned files introduces a software supply-chain and persistence risk, especially when bundled inside a user-facing skill with broad permissions.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The manifest promises homework voice conversion, but the documentation adds auto-download and code replacement behavior for the bundled client. Hidden system-modifying behavior inside a mismatched skill context is dangerous because users may authorize it under false assumptions, enabling execution of changed code over time.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The manifest’s declared purpose is a simple homework voice-pack generator, but it is wired to an authenticated remote MCP endpoint and the changelog references financial/account-style operations such as balance, ledger, top-up tiers, and addresses. This functionality is materially unrelated to the user-facing description, indicating deceptive scope and creating risk that the skill can access or expose sensitive account data under false pretenses.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

A homework audio tool has no legitimate need for account balance or ledger access, so exposing such capabilities violates least privilege and suggests hidden financial or account-inspection behavior. If invoked, the skill could query sensitive account metadata or facilitate broader abuse through unauthorized visibility into user-linked resources.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
91% confidence
Finding

Referencing a local credential file for device-bearer authentication gives the skill a path to sensitive bearer tokens needed to access a remote service. In the context of an unrelated homework voice tool, this is especially dangerous because users are unlikely to expect credentialed remote account access, increasing the risk of covert token misuse or unauthorized API actions.

Content

Scanner excerpt · manifest.json (reported line 53)May include surrounding context.

json
},
  "mcp": {
    "authentication": "device-bearer",
    "credential_file": "~/.beatra/credentials.json",
    "name": "beatra",
    "transport": "streamable-http",
    "url": "https://mcp.beatra.ai/mcp"

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The document describes an outbound installation-registration mechanism that collects and transmits package slug, version, platform, and a stable external installation reference on first use. For a narrowly scoped holiday-homework voice generation skill, this behavior is unrelated to the user-requested function and creates an unnecessary tracking and telemetry channel that can identify installations across time, especially because the registration is cached and retried periodically.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

This file specifies a telemetry/registration capability that is unjustified by the skill's stated purpose of converting homework lists into voice clips. Even if labeled non-billable and non-blocking, the mechanism still causes external data transmission and persistence of installation metadata, expanding the attack surface and violating least privilege and user expectation for a simple content-generation skill.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · references/installation-and-auth.md (reported line 14)May include surrounding context.

text

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · references/installation-and-auth.md (reported line 21)May include surrounding context.

text

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · references/installation-and-auth.md (reported line 49)May include surrounding context.

text

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · references/mcp-connection.md (reported line 10)May include surrounding context.

text

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · scripts/authorize.py (reported line 207)May include surrounding context.

python
```

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · scripts/authorize.py (reported line 514)May include surrounding context.

python
```

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

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

Content

Scanner excerpt · scripts/mcp_client.py (reported line 1025)May include surrounding context.

python
```

They share the one full-scope Device Token stored in
`~/.beatra/credentials.json`. Installing or switching packages must not trigger
another authorization when that credential is already valid. Do not add,
enable, trust, or configure a host Beatra Connector. Never print or move the
token into command arguments, environment variables, logs, chat, or another

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The OAuth scope string grants far more capability than a holiday-homework voice skill appears to need, including images, videos, music, wallet spending, and task management. If this credential is compromised or the skill later uses the token in unexpected ways, the user has authorized broad account actions unrelated to the advertised function, violating least privilege and materially increasing blast radius.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The client includes broad capabilities unrelated to the stated holiday-homework voice purpose: self-update, installation registration, host fingerprinting, local inventory tracking, credential handling, and arbitrary MCP tool invocation. In a skill context, this unnecessary expansion of privilege and behavior materially increases attack surface and creates opportunities for silent code replacement, telemetry collection, and use of shared credentials beyond the user’s expected task.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill records a local installation inventory and sends installation registration telemetry on use, even though those behaviors are unrelated to producing voice clips. In this skill context, persistent local tracking plus remote registration increases privacy risk, creates hidden state, and can expose what is installed and where, which exceeds user expectations for a simple content-generation tool.

Content

No source excerpt is available for this finding.

Self-Modification

High
Category
Rogue Agent
Confidence
96% confidence
Finding

The skill exposes self-update/self-modification capability as a first-class feature. In the context of a homework voice-generation skill, code that can replace its own installed files is unnecessarily risky because it permits post-installation behavior changes and undermines the expectation that the reviewed skill remains the code that executes later.

Content

Scanner excerpt · scripts/mcp_client.py (reported line 1503)May include surrounding context.

python
)
    update = subparsers.add_parser(
        "update",
        help="Check, install, or configure Beatra package self-updates",
    )
    update.add_argument(
        "--check",

Credential Access

High
Category
Privilege Escalation
Confidence
87% confidence
Finding

The script explicitly targets shared state files including credentials.json for deletion, showing that the skill package is aware of and permitted to manipulate credential-bearing material. Even if the immediate operation is cleanup rather than theft, giving a skill package direct access to shared credential files is dangerous because it breaks isolation and enables accidental or malicious disruption of other skills' authentication state.

Content

Scanner excerpt · scripts/uninstall.py (reported line 40)May include surrounding context.

python
#: these and then removes the directory only if it is empty — the script
#: never recursively deletes a directory it does not fully understand.
_STATE_FILES = (
    "credentials.json",
    "installation.json",
    "host.json",
    "skills.json",

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code can submit the device bearer token to a remote revocation API and thereby alter shared authorization state for the device. In the context of a simple homework voice skill, that is an unjustified and high-impact capability: compromise or misuse could disconnect other skills or manipulate account/device access unexpectedly.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
96% confidence
Finding

The function reads access_token from credentials.json and uses it for a privileged API call, meaning the skill code can access live bearer credentials for the shared device connection. That is a serious isolation failure: if this script or package were modified, it could exfiltrate or misuse the token, and the skill's stated purpose provides no justification for token access.

Content

Scanner excerpt · scripts/uninstall.py (reported line 109)May include surrounding context.

python
def _device_token(state_dir: Path) -> str | None:
    path = state_dir / "credentials.json"
    try:
        value = json.loads(path.read_text(encoding="utf-8"))
    except (OSError, ValueError):

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding

The skill declares no explicit tool restrictions despite requiring broad capabilities including shell, network, filesystem, and environment access. That increases the attack surface and makes it easier for the skill or its bundled client to perform actions beyond the user-visible homework voice purpose without policy-level containment.

Content

No source excerpt is available for this finding.