Back to skill

Security audit

Workspace Git Setup

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly coherent for local Git setup, but users should review it because the README suggests running an unpinned remote script and the default setup can commit every unignored file in the workspace.

Install from a pinned repository revision or audited package rather than the README's mutable main-branch curl command. Run --dry-run or --audit first, confirm the workspace path, and inspect which files would be committed before allowing the default setup to create a commit.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Warning
Location
README.md:35
Finding

Execution of an Unpinned Remote Script

Content
View full analysis
Remediation
View remediation
/skills/workspace-git-setup/scripts/setup.sh -o setup.sh ``` 2. Publish a SHA-256 checksum through a separately authenticated release channel and require verification before execution: ```bash echo " setup.sh" | sha256sum -c - bash setup.sh ``` 3. Prefer signed release artifacts or Git tags whose signatures users can verify. 4. Separate retrieval from execution and instruct users to inspect the downloaded script before running it. Do not recommend direct `curl | bash` execution. 5. Apply the same hardened instructions to both the English section at lines 35-36 and the duplicated Chinese section at lines 143-144. 6. Keep the bundled, audited `scripts/setup.sh` as the preferred execution path after obtaining the repository from a pinned revision. ]]>
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 (5)

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/setup.sh (reported line 28)May include surrounding context.

sh
# ── Workspace Git Ignore · 通用安全实践 ──────────────────────────────────────

# ── 敏感信息 / 凭证(安全核心,绝不可入 git)──────────────────────────────
.env
.env.*
*.pem
secrets/

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill’s trigger text ends with a broad catch-all clause that activates on 'any other phrasing' related to version tracking, repo audit, or backup intent. In agent systems, overly broad activation can cause the skill to run in contexts the user did not explicitly intend, leading to unintended repository initialization, config changes, staging, or commits in the current directory or an environment-selected workspace.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The script's comments, prompts, status messages, and usage guidance are written in Chinese throughout, which effectively forces a specific language for interaction. There is no opt-in, locale selection, or justification that this skill is intended only for a Chinese-speaking environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script performs git add -A followed by an automatic commit, which can sweep in all files in the workspace, including files the user did not intend to version. Although the script adds a protective .gitignore and offers audit/dry-run modes, the default setup path still makes broad repository changes in one step, increasing the risk of accidental source, config, or residual secret exposure.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The script reads GIT_AUTHOR_EMAIL from the environment and immediately prints the email address to stdout. This is a form of sensitive data disclosure in a code file, and there is no warning or masking despite the value being personally identifiable information that may end up in logs or terminal recordings.

Content

No source excerpt is available for this finding.