Back to skill

Security audit

ovulation-tracking

Security checks for vulnerabilities and agentic risk

Overview

This fertility tracking skill matches its general purpose, but it handles highly sensitive health data with misleading privacy claims, automatic partner-email behavior, plaintext storage, and unpinned install instructions.

Install only after reviewing the code and configuration carefully. Treat the Oura token, cycle history, symptoms, LH results, and partner email settings as highly sensitive; avoid enabling partner/email alerts unless every recipient has explicitly consented, store secrets outside shared or backed-up folders, and do not rely on predictions as medical advice or contraception guidance.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (3)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:46
Finding

Mutable Remote Package Execution During Installation

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
v2/index.js:306
Finding

Sensitive Reproductive Health Records Stored as Unencrypted JSON

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
config.example.txt:5
Finding

Reusable Email Credential Stored in Plaintext Configuration

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (38)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The skill claims broad multi-signal fusion and learning, but the finding suggests narrower temperature-centric logic and only partial LH usage. This discrepancy is especially concerning in a reproductive-health skill because inaccurate user expectations can drive unsafe reliance and unnecessary disclosure of sensitive data.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The skill claims broad multi-signal fusion and learning, but the finding suggests narrower temperature-centric logic and only partial LH usage. This discrepancy is especially concerning in a reproductive-health skill because inaccurate user expectations can drive unsafe reliance and unnecessary disclosure of sensitive data.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
86% confidence
Finding

The skill claims broad multi-signal fusion and learning, but the finding suggests narrower temperature-centric logic and only partial LH usage. This discrepancy is especially concerning in a reproductive-health skill because inaccurate user expectations can drive unsafe reliance and unnecessary disclosure of sensitive data.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill claims broad multi-signal fusion and learning, but the finding suggests narrower temperature-centric logic and only partial LH usage. This discrepancy is especially concerning in a reproductive-health skill because inaccurate user expectations can drive unsafe reliance and unnecessary disclosure of sensitive data.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The privacy section says data never leaves the user's machine, yet elsewhere the skill documents partner email alerts, which necessarily transmit sensitive fertility information externally. False privacy assurances around reproductive health data can cause users to enable features they would otherwise reject, resulting in unauthorized disclosure.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill transmits sensitive health data to the Oura API using an authorization token, but the file shows no user-facing disclosure, consent flow, or minimization controls. Because fertility and readiness data are highly sensitive, undisclosed external transmission can expose intimate health information and create compliance and privacy risks.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

These emails contain explicit fertility status, ovulation timing, and sexual activity recommendations, which are highly sensitive medical and reproductive data. Sending them without explicit warning, confirmation, and privacy safeguards risks disclosure to unintended recipients through shared mailboxes, forwarding, or account compromise.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The health-status alert emails disclose inferred illness or severe stress indicators derived from wearable data without prior disclosure or confirmation. Such unsolicited transmission of sensitive health assessments can cause privacy harm and may expose medical inferences to anyone with access to the email account.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The README promotes fertility and ovulation predictions with quantified accuracy claims and personalized learning, but it does not prominently warn that the tool is not medical advice and should not be used as the sole basis for health, conception, or contraception decisions. In this context, users may rely on the predictions for reproductive or medical decisions, creating meaningful risk of harm from false confidence, delayed care, or unintended pregnancy.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The configuration example includes partner email alerts and automated daily checks for highly sensitive reproductive data, but it does not clearly warn users that intimate fertility information may be transmitted to another person or over external channels. Because fertility status, symptoms, and cycle predictions are sensitive health-related data, accidental or uninformed disclosure can cause privacy harm, relationship harm, or coercive misuse.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
94% confidence
Finding

This code produces ovulation predictions and confidence values from highly sensitive reproductive health signals, but it contains no guardrails, uncertainty messaging, or indication that outputs are not medical advice. In a fertility-tracking context, users may reasonably rely on these predictions for conception or pregnancy-avoidance decisions, so inaccurate or overconfident output can cause significant real-world harm.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The prediction logic returns actionable, confident-sounding fertility guidance such as 'Optimal TTC timing' and likely ovulation windows without any warning that the output is probabilistic, fallible, and unsuitable as a sole basis for contraception or conception planning. Because this is a reproductive-health use case, omission of uncertainty and safety disclaimers materially increases the risk of harmful real-world decisions, including unintended pregnancy or delayed care.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The changelog claims the feature was 'tested with real user data' but gives no indication that consent, de-identification, retention limits, or privacy protections were in place. Because the skill handles reproductive and health-related signals, use of real user data without documented safeguards raises serious privacy and compliance concerns and may normalize unsafe data practices.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The changelog describes automated partner email alerts and timing guidance tied to LH surge and fertility status, but provides no indication of explicit consent, recipient verification, opt-in controls, or warnings about sharing highly sensitive reproductive health data. In a fertility-tracking context, this can expose intimate medical information to third parties, creating substantial privacy, safety, and relationship-abuse risks if misconfigured or enabled without clear user approval.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The feature list states 'LH test strip analysis: Photo-based progression tracking' as an available capability. However, the roadmap and contributing sections explicitly list 'LH strip image recognition' as future work, which contradicts the earlier claim that the agent currently analyzes strip photos.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The README understates or omits clear warnings that highly sensitive reproductive health data may be shared with third-party services like Oura, Open-Meteo, SMTP providers, and with a partner via email notifications. Because fertility status and cycle data are especially sensitive, inadequate disclosure can lead to privacy harm, unintended disclosure, and compliance issues.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
89% confidence
Finding

The README instructs users to store a long-lived Oura access token in a plaintext file under the home directory, which increases the chance of credential theft through local compromise, backups, logs, or overly broad file permissions. Since this token grants access to health-related wearable data, compromise can expose sensitive biometric and cycle information.

Content

Scanner excerpt · README.md (reported line 67)May include surrounding context.

Visit Oura Cloud

https://cloud.ouraring.com/personal-access-tokens

Create token, save to file

echo "YOUR_TOKEN_HERE" > ~/.config/oura/token.txt

text

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The trigger phrase 'My period started today' is natural language that could easily appear in ordinary conversation, journaling, or health discussions, causing unintended activation. In this skill, accidental invocation is more sensitive because it can initiate fertility tracking workflows and partner notifications involving intimate health data.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The LH Testing Protocol tells users 'Agent analyzes and tracks progression' after they send a strip photo. Elsewhere, the document lists LH strip image recognition as a not-yet-built feature, so this operational instruction overstates what the skill currently does.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding

The skill documents capabilities that require network access and local secret/config handling, but it declares no explicit tool scope or permissions. In a health-related skill handling reproductive data, missing permission boundaries increases the chance of overbroad access, hidden exfiltration, or unsafe execution by an agent runtime.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
94% confidence
Finding

Using npx clawhub@latest install fertility-tracker relies on a moving package version, which can change over time and pull unreviewed code into the installation path. This creates a supply-chain risk where a compromised or malicious upstream release could execute during install.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The configuration stores an Oura API token and partner email settings, but the documentation does not clearly warn users that the skill accesses sensitive health data and may transmit it through integrations. Inadequate disclosure around secret handling and health-data sharing creates elevated privacy and account-security risk.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill describes automatic partner email alerts for LH surge and ovulation timing without a clear privacy warning or consent framing. These messages contain highly sensitive reproductive health information, and automatic transmission can expose intimate data through inbox compromise, shared accounts, forwarding, or mistaken recipients.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

Claiming that all analysis happens locally conflicts with documented use of the Oura API, which requires external data exchange. Even if computation is local after retrieval, the statement is materially misleading about the privacy model and third-party exposure.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
94% confidence
Finding

The repeated npx clawhub@latest command again exposes users to unpinned remote code execution during installation. Repetition in prominent call-to-action sections increases the chance that users will follow the unsafe path without review.

Content

No source excerpt is available for this finding.