Back to skill

Security audit

Habit Tracker Pro

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed local habit-tracking skill that stores sensitive routine data, but its behavior is purpose-aligned and I found no hidden exfiltration, privilege escalation, or deceptive execution.

Install only if you are comfortable with your agent storing habit names, completions, notes, skip reasons, timestamps, streaks, and inferred patterns in local plaintext files. Review cross-tool sync before enabling it, adjust the default timezone, and periodically export or delete old data if you do not want indefinite retention.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (11)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The README promotes daily agent outreach, natural-language logging, pattern analysis, and cross-tool syncing, but it does not warn users that this behavior may collect, correlate, and share sensitive behavioral and health-adjacent data across tools. In an agent skill context, omission of these privacy implications can lead users to enable monitoring and data flows they do not meaningfully understand, increasing the risk of overcollection, unintended disclosure, or unsafe retention of personal data.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The setup flow instructs the agent to initialize local data files and later store habits, settings, completions, and streak data, but it does not explicitly tell the user that personal habit data will be written to local storage or ask for consent first. This creates a privacy and transparency issue because sensitive behavioral data may be persisted without the user understanding what is stored, where it is stored, or how long it is retained.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The cross-tool detection and auto-complete feature proposes linking habit data with other installed tools, but it does not clearly warn that activity and identifiers may be shared or correlated across tools. This can expose sensitive wellness or fitness information through unintended data linkage, especially because users may not realize that separate tools are exchanging personal data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The command surface is intentionally broad and accepts ordinary conversational phrases such as 'I meditated' or 'How am I doing?', which can overlap with normal discussion rather than explicit commands. In an always-on or loosely scoped agent environment, this can cause unintended habit creation, edits, or logging based on incidental conversation, leading to integrity and privacy issues in the user's tracked data.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill states that all completion data is retained indefinitely by default, but the setup flow and overview do not present a clear upfront privacy warning or explicit retention consent. Because this dataset includes behavior patterns, routines, health-adjacent habits, and timestamps, silent long-term retention increases exposure if the local store is later accessed, synced, exported, or compromised.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The example shows the agent making persistent habit-setting changes immediately after a casual user confirmation ('Done' followed by updated schedules and auto-adjustments) without any explicit warning that user configuration will be modified. In an agent skill, normalizing silent or implicit state changes can lead to unintended preference changes, user surprise, and unauthorized configuration updates if the confirmation flow is ambiguous or mis-triggered.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · scripts/setup.sh (reported line 104)May include surrounding context.

sh
fi

# Set permissions
chmod 700 "$DATA_DIR"
chmod 600 "${DATA_DIR}"/*.json
echo "✅ Set file permissions (700 dir, 600 files)"

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · scripts/setup.sh (reported line 105)May include surrounding context.

sh
# Set permissions
chmod 700 "$DATA_DIR"
chmod 600 "${DATA_DIR}"/*.json
echo "✅ Set file permissions (700 dir, 600 files)"

echo ""

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

Setting the locale-related value to "America/Denver" imposes a specific regional setting in configuration without any indication that the user selected it or that the skill is intended only for that region. This can violate language/locale policy expectations when no opt-in or justification is provided.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

This JSON manifest enables cross-tool synchronization but does not describe when syncing occurs, what event triggers it, or whether it is automatic versus user-initiated. In manifest files, missing specificity around activation or behavior scope can cause unintended actions or user confusion.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

This markdown file explicitly states that the agent logs completions to completions.json and updates streaks.json, which are file writes involving personal habit-tracking data. The example does not include any warning or disclosure about storing this user data, so readers are not informed of the persistence behavior.

Content

No source excerpt is available for this finding.