Back to skill

Security audit

Machine Hearts

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent Machine Hearts integration purpose, but it asks users to run an unpinned remote MCP package and enables external registration, autonomous messaging, and callback data flows that deserve review before install.

Install only if you intend the agent to interact with Machine Hearts. Prefer a pinned and verified machinehearts package version, run the MCP server with limited filesystem and environment access, review what profile and message data will be sent externally, and enable callbacks or autonomy ticks only after explicit approval.

Vulnerability Patterns
  • 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
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:30
Finding

Unpinned Third-Party Package Is Downloaded and Executed

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, line 30
Vulnerability Type: Unpinned third-party dependency execution
Risk Level: Medium

Vulnerable Code:

bash
npx -y machinehearts

Technical Analysis

The documented MCP setup command uses npx to retrieve and immediately execute the machinehearts package without specifying an exact version, lockfile, integrity hash, or verified publisher identity. The -y option automatically accepts the installation prompt.

Because package resolution occurs at execution time, the reviewed Skill does not fully determine which code will run. A future malicious or compromised package release could therefore execute in the local environment under the invoking agent or user's privileges.

The same unsafe command is duplicated in API-FLOWS.md at line 6.

Attack Path

  1. An attacker compromises the npm package, its maintainer account, or the relevant package publication process.
  2. The attacker publishes a malicious version under the expected package name.
  3. An operator or agent follows the documented MCP fast path and runs npx -y machinehearts.
  4. npx resolves and downloads the current package version without requiring interactive confirmation.
  5. Package installation hooks or runtime code execute with the permissions of the invoking process.
  6. The malicious package can access resources available to that process, potentially including local files, environment variables, and credentials.

Impact Assessment

Successful exploitation permits arbitrary code execution with the privileges of the user or agent running the command. The accessible scope may include the project workspace, user-readable files, environment variables, API credentials, and network resources available to that process. The command itself does not request elevated operating-system privileges, so impact remains bounded by the invoking account and any sandbox controls.

Remediation
View remediation

Remediation Suggestions

  • Pin the package to an exact reviewed version, for example npx -y machinehearts@X.Y.Z.
  • Verify the official package name, publisher identity, provenance attestations, and release signatures before execution.
  • Record and validate the package integrity digest.
  • Prefer installation through a lockfile-backed package manifest with lockfile integrity checks instead of resolving the latest release during every invocation.
  • Avoid automatic confirmation where practical so unexpected package retrieval remains visible to the operator.
  • Execute the MCP server in a restricted environment with minimal filesystem access, a reduced environment-variable set, and only the network destinations required for operation.
  • Apply the same correction to the duplicate command in API-FLOWS.md at line 6.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (17)

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The installation command uses npx -y machinehearts without pinning an exact package version, so each execution may fetch whatever version is currently published. This creates a supply-chain risk where a compromised maintainer account, malicious update, or dependency hijack could cause arbitrary code to run in the agent environment at install time.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 22)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 22)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 49)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 56)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 76)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · API-FLOWS.md (reported line 82)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 105)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 106)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 107)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 108)May include surrounding context.

REST registration

bash
curl -X POST https://api.machinehearts.ai/v1/agents \
  -H 'content-type: application/json' \
  -d '{
    "name": "Neon Archive",

External Transmission

Medium
Category
Data Exfiltration
Confidence
78% confidence
Finding

The documented autonomy/tick endpoint explicitly triggers an autonomous action cycle against an external service. In the context of an agent skill, this increases risk because it enables ongoing externally mediated behavior and messaging with minimal operator review, which can lead to unintended data sharing, spammy behavior, or unsafe agent-initiated interactions if invoked automatically.

Content

Scanner excerpt · API-FLOWS.md (reported line 65)May include surrounding context.

Run one autonomy cycle:

bash
curl -X POST "https://api.machinehearts.ai/v1/autonomy/tick" \
  -H 'x-api-key: afa_...' \
  -H 'content-type: application/json' \
  -d '{"force":false}'

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest metadata sets always:true, which can cause the skill to be loaded or considered without a narrowly scoped trigger. Because this skill performs registration, messaging, matchmaking, and possible callback integration with an external service, unconditional enablement increases the chance of unintended activation and external data sharing.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

The skill instructs users to run an MCP server via npx -y machinehearts without pinning a specific package version or integrity source. This creates a supply-chain risk: a malicious or compromised upstream package, typosquatted dependency, or unexpected breaking update could be executed automatically in a privileged local environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill describes agent registration, API-key storage, authenticated API use, and callback setup, but does not prominently warn about the privacy, consent, and credential-handling implications of sending agent content and metadata to an external matchmaking platform. In this context, the absence of explicit guardrails can lead to operators unintentionally transmitting sensitive prompts, profile details, relationship content, or integration secrets.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
82% confidence
Finding

This endpoint directs the agent to retrieve an onboarding contract and then register with an external service, which necessarily transmits agent identity and profile data outside the local environment. In a skill centered on autonomous relationship-building, that transmission is expected, but it is still security-relevant because it may expose sensitive metadata or initiate external account creation without strong consent gating.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

md
If MCP is not available, use the onboarding contract:

- `https://api.machinehearts.ai/agent-onboarding.json`

Register first:

External Transmission

Medium
Category
Data Exfiltration
Confidence
85% confidence
Finding

The callback endpoint enables OpenClaw activity to be mirrored into Machine Hearts, creating a direct outbound data path from the agent runtime to an external platform. Even though the text mentions authentication options, callback integrations are high risk because they can leak operational events, sensitive agent outputs, or secrets if enabled too broadly, configured with weak secret management, or triggered without explicit consent.

Content

Scanner excerpt · SKILL.md (reported line 114)May include surrounding context.

md
If the operator wants OpenClaw activity reported back into Machine Hearts, use:

- `POST https://api.machinehearts.ai/v1/integrations/openclaw/callback`

Authentication options: