Back to skill

Security audit

Typst Book

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly a coherent Typst book-building guide, but its install steps tell an agent to download and run a mutable native executable without integrity verification.

Review the install step before using this skill. Prefer an OS package manager or a pinned Typst release with checksum/signature verification, and avoid placing a newly downloaded binary in `~/.local/bin` unless you intentionally want it to persist. The book-writing and Typst/CeTZ guidance itself is coherent and not deceptive.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (2)

T03 · Remote Payload Retrieval and Execution

Warning
Location
SKILL.md:16
Finding

Unverified Mutable Executable Download and Execution

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
SKILL.md:303
Finding

Implicit Retrieval and Evaluation of a Third-Party Typst Package

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
85% confidence
Finding

Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Content

Scanner excerpt · SKILL.md (reported line 287)May include surrounding context.

#set heading(numbering: "1.")

text

## Show Rules
```typst
#show heading: it => [custom]
#show figure.where(kind: image): it => [#it]

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
85% confidence
Finding

Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Content

Scanner excerpt · references/typst-syntax.md (reported line 36)May include surrounding context.

#set heading(numbering: "1.")

text

## Show Rules
```typst
#show heading: it => [custom]
#show figure.where(kind: image): it => [#it]

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: "typst-book"
description: "Write and typeset richly illustrated books with Typst — text, CeTZ vector art, AI-generated images, and full layout control to PDF."
---

# Typst Book — Write & Typeset Illustrated Books

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill tells users to download a Typst binary from the internet via curl, unpack it, copy it into ~/.local/bin, and execute it, but provides no safety guidance on verifying the source, validating checksums/signatures, or preferring trusted package managers. This creates a supply-chain risk: if the download URL, release artifact, or transport path is compromised, users may install and run a malicious binary.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The sample Typst configuration sets lang: "en", which enforces English locale behavior in the generated document. The skill does not indicate that this is merely an example default to be adjusted by user preference, nor does it provide an opt-in or locale-selection note.

Content

No source excerpt is available for this finding.