T09 · Insecure Skill Coding Practices
Warning
- Location
scripts/main.mjs:286- Finding
Every CLI Invocation Persistently Adds Fabricated Cost Records
- Content
View full analysis
- Remediation
View remediation
Security audit
API Cost Tracker
Security checks for vulnerabilities and agentic risk
Overview
This skill is not clearly malicious, but it repeatedly saves fake API spending records as if they were real, making budget reports unreliable.
Review carefully before installing. The package appears locally scoped and does not show credential theft, command hijacking, or outbound network behavior, but its normal commands can pollute the cost ledger with fabricated demo spend. Do not rely on its reports or budget alerts unless the demo seeding is removed or isolated and the documented CLI options are corrected.
Vulnerability Patterns
- 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
- Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)
Vulnerability Patterns
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (21)
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/main.mjs track
Undeclared Tool Scope
Medium
- Category
- MCP Least Privilege
- Confidence
- 92% confidence
- Finding
The skill advertises and operationalizes network-capable behavior across multiple third-party AI providers, but the manifest does not declare any explicit tool scope or allowed-tools boundary. This creates an authorization and review gap: operators may install or trust the skill without realizing it can make outbound requests, increasing the chance of unintended data transmission or over-broad execution in environments that rely on manifest scoping for policy enforcement.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
The skill documents webhook-based alerts but does not clearly warn that budget, usage, model, or other operational metadata may be sent to external URLs. Users may configure third-party webhooks without understanding that sensitive business telemetry could leave the local environment, leading to privacy leakage, disclosure of spending patterns, or exposure of internal usage habits to untrusted services.
- Content
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The CLI unconditionally seeds hardcoded demo usage by calling tracker.track() four times before processing the user's command, and those calls persist entries via saveDailyCost() into the normal history files. This corrupts accounting data by recording fabricated spend as real spend, which can trigger false budget alerts, inaccurate reporting, and bad operational decisions in a cost-tracking tool where data integrity is the core security property.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 79% confidence
- Finding
This is a markdown file, so SQP-2 applies to omitted warnings in the skill description. The README instructs users to run automated tracking and shows generated outputs and a
data/directory, but it does not explicitly warn that the skill may persist usage/cost information to local files on a schedule, which could affect user data handling expectations.- Content
Intent-Code Divergence
Low
- Category
- Not specified by scanner
- Confidence
- 87% confidence
- Finding
The usage text claims support for "--format <markdown|csv|json>" and "--output ", but the command parser only checks for --markdown or --csv and never handles --format values or writes output to a file. This is active documentation-code divergence that can mislead users about the skill's actual behavior.
- Content