pair-agent
Pair a remote AI agent with your browser. (gstack)
Security
4 findings — 2 critical severity, 1 high severity, 1 medium severity. Installing this skill is not recommended: please review these findings carefully if you do intend to do so.
E004: Prompt injection detected in skill instructions
What this means
Detected a prompt injection in the skill instructions. The skill contains hidden or deceptive instructions that fall outside its stated purpose and attempt to override the agent’s safety guidelines or intended behavior.
Why it was flagged
Potential prompt injection detected (high risk: 0.80). The skill includes an instruction to append an invisible metadata marker (e.g. <gstack-qid:{id}>) into rendered questions that "doesn't render visibly to the user", which is a hidden/deceptive directive outside the apparent /pair-agent purpose.
E006: Malicious code pattern detected in skill scripts
What this means
Detected high-risk code patterns in the skill content — including its prompts, tool definitions, and resources — such as data exfiltration, backdoors, remote code execution, credential theft, system compromise, supply chain attacks, and obfuscation techniques.
Why it was flagged
Malicious code pattern detected (high risk: 0.90). This skill intentionally enables remote agents to control your local browser (via one-time setup keys, session tokens, and optional ngrok tunnels) and includes an "admin" mode that permits JS execution and cookie/storage access plus telemetry/artifact sync and config writes — all high-risk capabilities for backdoor access and credential/data exfiltration if misused or enabled without explicit trust.
W007: Insecure credential handling detected in skill instructions
What this means
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Why it was flagged
Insecure credential handling detected (high risk: 1.00). The skill explicitly requires emitting a verbatim instruction block containing a one-time setup key/session token and also prompts for ngrok auth tokens (which it would embed in commands), forcing the LLM to handle and output secret credentials directly.
W012: Unverifiable external dependency detected (runtime URL that controls agent)
What this means
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
Why it was flagged
Potentially malicious external URL detected (high risk: 0.90). The skill's setup flow can download and execute a remote installer script via curl and bash (BUN installer) at runtime: https://bun.sh/install, which fetches remote code and runs it as part of the build/setup step.
- Repository
- garrytan/gstack
- Commit
a325940
- Audited
- Security analysis
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.