Security audit

Git基础版

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward Git helper skill with expected command execution, but users should be careful with state-changing and remote Git actions.

Install this only if you want an agent to help run Git commands. Review generated commands before commits, merges, pulls, or pushes, confirm the remote URL and target branch, and avoid passing untrusted branch names, file paths, or commit messages without checking how your agent executes shell commands.

A.I.G

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)

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:77
Finding: Shell Command Injection Through Unvalidated Git Parameters
Content: View full analysis
` or `git commit -m ""`. 3. The Agent passes the generated string to a shell through its unrestricted `e ...[truncated 1115 chars]
Remediation: View remediation
`, to prevent option injection. 7. Preserve commit messages as a single opaque process argument rather than manually adding quotation marks. 8. Resolve and validate paths against the intended repository root before execution. 9. Require explicit user confirmation before `push`, branch deletion, destructive resets, forced updates, or operations affecting a remote repository. 10. Restrict execution to the intended repository directory and apply platform-level filesystem and network sandboxing. 11. Avoid returning command output that may contain credentials, authenticated remote URLs, tokens, or other sensitive environment details. ]]>

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The remote synchronization feature covers clone, fetch, pull, and push but provides no warning that these operations transmit repository data and credentials over the network or can change local and remote state. In an agent-enabled environment, this increases the risk of accidental data exfiltration, pushing to the wrong remote, or interacting with untrusted repositories without adequate user awareness.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The skill explicitly instructs the agent to execute Git commands and capture output, but it does not require confirmation, dry-run behavior, or warnings before repository-modifying operations such as add, commit, merge, pull, and push. In an agent context with exec capability, this can cause unintended local history changes, merges, or remote updates if user intent is ambiguous or inputs are interpreted incorrectly.
Content: No source excerpt is available for this finding.