ClawHub
Back to skill

Security audit

Git Manager

Security checks for vulnerabilities and agentic risk

Overview

This skill gives Git maintenance guidance and clearly warns before destructive Git operations.

Install only if you want an agent to help with advanced Git maintenance. Review every suggested destructive command carefully, verify the branch and commit, and keep a backup branch, tag, stash, or patch before approving changes that rewrite history or affect remotes.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Tool Parameter Abuse

High
Category
Tool Misuse
Content
## Quick start
1. Trigger the skill with a command like `git manager` or `bisect issue` inside a repository.
2. Follow the prompts for diagnostics, branch cleanup, stash review, or log analysis.
3. Execute the suggested commands and confirm before destructive operations such as `git branch -D`, `git reset --hard`, `git clean -fd`, or `git push --force` (the skill prefers `git push --force-with-lease` unless the user explicitly authorizes raw `--force`).

Every destructive recommendation uses a two-step confirmation protocol: show the current branch and HEAD commit/tag, outline a backup action (tag, temporary branch, stash/export), and then require an explicit `YES` before running `branch -D`, `reset --hard`, `git clean -fd`, or any forced push. This reinforces the safeguards embedded in the skill's workflow.
Confidence
70% confidence
Finding
Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Tool Parameter Abuse

High
Category
Tool Misuse
Content
## Quick start
1. Trigger the skill with a command like `git manager` or `bisect issue` inside a repository.
2. Follow the prompts for diagnostics, branch cleanup, stash review, or log analysis.
3. Execute the suggested commands and confirm before destructive operations such as `git branch -D`, `git reset --hard`, `git clean -fd`, or `git push --force` (the skill prefers `git push --force-with-lease` unless the user explicitly authorizes raw `--force`).

Every destructive recommendation uses a two-step confirmation protocol: show the current branch and HEAD commit/tag, outline a backup action (tag, temporary branch, stash/export), and then require an explicit `YES` before running `branch -D`, `reset --hard`, `git clean -fd`, or any forced push. This reinforces the safeguards embedded in the skill's workflow.
Confidence
65% confidence
Finding
Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Tool Parameter Abuse

High
Category
Tool Misuse
Content
3. **Branch cleanup** – use `git branch --merged` versus `git branch --no-merged` to find stale branches. Recommend `git branch -d <branch>` for merged work and `git branch -D` only after re-confirming the target branch via a safety prompt to avoid deleting active work.
4. **Stash management** – suggest `git stash list`/`git stash show` to catalogue hidden work. Encourage naming stashes with `git stash push -m "description"` and verify the exact entries before `git stash drop`/`pop`, reminding the user to keep a copy (`git stash branch <name>`) if they need extra safety.
5. **Log analysis** – guide through `git log --oneline --graph --decorate`, `git reflog`, and `git log @{u}` to understand recent operations. Offer commands like `git show <commit>` for inspection and highlight the importance of reviewing commit messages before reverting or cherry-picking.
6. **Safety prompts** – before destructive commands (e.g., `reset --hard`, `git clean -fd`, branch deletion, `rebase`, `push --force`), run the two-step destructive confirmation protocol: (1) display the current branch name and HEAD commit hash or tag, verify the target commit, and remind the user to take a backup action (tag, temporary branch, stash, export patch, etc.); (2) require an explicit textual `YES` reply before running the command. Always pair the recommendation with `git status`, `git log -1`, or a `git tag` snapshot so the user can see exactly what would change, and prefer `git push --force-with-lease` over `--force` unless the situation explicitly warrants the risk.

## Deliverables
- Provide step-by-step command sets for diagnostics (bisect, log review, stash recovery).
Confidence
70% confidence
Finding
Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Static analysis

No suspicious patterns detected.