Back to skill

Security audit

Construction Progress Report Generator

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only construction reporting assistant that structures user-provided project information and does not include code, persistence, data exfiltration, or automatic sending behavior.

Safe to install for construction progress reporting. Review reports before sharing them, avoid unnecessary personal contact details, and be careful with commercial disputes, payment amounts, claims, and site photos intended for external stakeholders.

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
  • 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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The activation examples include generic phrases such as asking for help with a project report or progress update, which could cause the skill to trigger in broader workplace conversations than intended. Over-broad activation increases the chance that unrelated or sensitive business discussions are pulled into this reporting workflow and transformed into a structured stakeholder-facing document.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill explicitly collects names/roles for a distribution list and prepares content for clients and management, but provides no privacy or confidentiality guardrails. This can lead users to disclose personal data, commercially sensitive project information, dispute details, or stakeholder contact information without warning or minimization, increasing risk of inappropriate sharing or retention.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 321)May include surrounding context.

md
1. **Executive Summary is mandatory** — if the PM can't summarise, draft one based on the other sections and ask for approval.
2. **Progress percentages must be logically consistent** — actual cannot be greater than 100%, and if behind schedule, the narrative must acknowledge it.
3. **Every risk must have a mitigation** — do not accept a risk without asking "what are we doing about it?"
4. **Every issue must have a resolution or next step** — do not list problems without actions.
5. **Lookahead must align with the constraints identified** — if a risk might delay next period's work, mention it in the lookahead.
6. **Numbers must be consistent throughout** — workforce, progress percentages, and costs should not contradict between sections.