ClawHub
Back to skill

Security audit

JIRA

Security checks across malware telemetry and agentic risk

Overview

This Jira skill appears legitimate, but it gives agents broad, under-scoped ways to change live Jira data and handle API-token-based access.

Install only if you trust the publisher and intend agents to work with your Jira account. Use least-privilege Jira credentials, avoid giving broad project-admin tokens, require explicit confirmation before any create/update/move/assign/comment/sprint action, and prefer the scoped MCP/CLI path over raw curl commands with API tokens.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (6)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The document expands beyond Jira MCP usage into shell/REST fallback instructions, including direct CLI and curl examples. In an agent skill context, this can encourage tool or channel escape from the intended MCP boundary and normalize handling raw credentials and external commands, increasing the chance of unsafe execution or secret exposure.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger list includes very generic terms like "issue," "ticket," "sprint," and "backlog," which can cause the Jira skill to activate in conversations unrelated to Jira. In a skill that can perform write actions against an external system, unintended activation increases the chance of confusing context, exposing Jira data unnecessarily, or prompting users toward accidental modifications.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger description includes very generic terms like `issue` and `ticket`, which can activate the skill for unrelated conversations and cause the agent to enter a Jira-oriented workflow unexpectedly. That increases the chance of inappropriate tool use, unnecessary environment probing, or prompting the user for Jira actions when they did not intend to interact with Jira.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The reference documents multiple state-changing Jira operations such as create, move, assign, comment, sprint modification, and linking, but does not warn that these commands modify live remote project data. In an agent skill context, this increases the risk that a model or user treats examples as safe read-only actions and performs unintended writes to production Jira projects.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The documentation shows credential-based API access using JIRA_USER and JIRA_API_TOKEN without any safety guidance on storage, redaction, or execution boundaries. In a skill consumed by agents, this creates a realistic risk that secrets are requested, echoed, logged, or misused in generated commands.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
# Create with custom fields
jira issue create -tStory -s"Summary" --custom story-points=3

# Skip prompts for optional fields
jira issue create -tTask -s"Quick task" --no-input

# Open in browser after creation
Confidence
81% confidence
Finding
Skip prompt

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.