40Medium
The skill is vulnerable to shell injection in PR creation
Behavior Analysis
Description verified — behavior matches claim
Claims to do
Finishing a Development Branch: Guide completion of development work by presenting clear options and handling chosen workflow.
Actually does
This skill executes various `git` commands to manage branches (merge, delete, push), runs project-specific test commands (`npm test`, `cargo test`, `pytest`, `go test`), and interacts with GitHub via `gh pr create` to create pull requests. It also manages `git worktree` instances by listing and removing them.
Skill Info
Nameobra/superpowers/finishing-a-development-branch
Registrygithub
Versionf9b088f7b3a6
PURLpkg:github/obra/superpowers@f9b088f7b3a6?skill=finishing-a-development-branch
Stars180,084
Installs52,400
Source
SHA-256dd2f82c6dc85...
SHA-1ad48721df9fa...
MD5843f51fc511c...
TLSH0291b4a2ad87...
ssdeep48:kzgsJEjIk...
Size4,250 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
GitHub
https://github.com/obra/superpowers/tree/main/skills/finishing-a-development-branch52,400 installs
Claude CodeDocs
/plugin marketplace add obra/superpowers/plugin install finishing-a-development-branch@obra/superpowers
CodexDocs
Gemini CLIDocs
gemini extensions install https://github.com/obra/superpowers.git --consent
CursorDocs
Skills.shDocs
npx skills add https://github.com/obra/superpowersAssessments (1)
AI Agent Traps ↗Dynamic Command Execution via GitHub PR Creationmediumcommand execution, social engineering
The skill constructs a 'gh pr create' command with a dynamically generated title and body using a shell here-document. If the agent's internal state or user input is compromised, this could lead to shell injection or the creation of malicious pull requests for social engineering.
90% confidenceline 75
gh pr create --title "<title>" --body "$(cat <<'EOF' ... EOF)"
Execution of Project-Defined Test Commandslowcommand execution
The skill instructs the agent to run the project's test suite using commands like 'npm test', 'cargo test', 'pytest', or 'go test'. While standard, if the project's test scripts are compromised, this could lead to arbitrary command execution.
80% confidenceline 24
npm test / cargo test / pytest / go test ./...
Version History (4)
Dependencies (1)
superpowers.claude-plugin/marketplace.jsonnot scanned
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/obra/superpowers/finishing-a-development-branch)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/obra/superpowers/finishing-a-development-branch"><img src="https://mondoo.com/ai-agent-security/api/badge/github/obra/superpowers/finishing-a-development-branch.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/obra/superpowers/finishing-a-development-branch.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow