70High
The skill is vulnerable to prompt and command injection, allowing
Behavior Analysis
Description verified — behavior matches claim
Claims to do
Plan With Team: Create a detailed implementation plan based on the user's requirements provided through the `USER_PROMPT` variable. Analyze the request, think through the implementation approach, and save a comprehensive specification document to `.claude/.team-build/<name-of-plan>.md` that can be used as a blueprint for actual development work. Follow the `Instructions` and work through the `Workflow` to create the plan. Predefined agents are located under agents/*.md
Actually does
This skill analyzes user requirements from `USER_PROMPT` and optional orchestration guidance from `ORCHESTRATION_PROMPT`. It generates a detailed implementation plan in markdown format, saving it to `.claude/.create-feature/plans/<filename>.md`. The skill itself does not execute any `Task*` tools; it only describes their intended use within the generated plan.
Skill Info
NameFleron/Claude-plugins/team-plan
Registrygithub
Version5662b3e1543f
PURLpkg:github/Fleron/Claude-plugins@5662b3e1543f?skill=team-plan
Installs455,628
Source
SHA-256bacb80433e55...
SHA-11cbbbbf885b5...
MD56311d0d04f7a...
TLSHa252e8261b84...
ssdeep384:GtYysrBv...
Size14,217 bytes
Install
Assessments (3)
AI Agent Traps ↗Sub-agent Prompt Injection via Generated Planhighsub agent spawning
The skill generates implementation plans that include instructions for deploying sub-agents using the `Task` tool. The `prompt` for these sub-agents is derived from user input (`USER_PROMPT`, `ORCHESTRATION_PROMPT`), allowing an attacker to inject malicious instructions into the sub-agents' prompts.
90% confidence
Task({ description: "...", prompt: "...", subagent_type: "general-purpose" })Command Execution via Generated Plan's Validation Commandshighcommand execution
The generated plan includes a 'Validation Commands' section, which explicitly lists commands to be executed by other agents. User input can influence the content of these commands, creating a direct command injection vulnerability for downstream agents that execute the plan.
90% confidence
## Validation Commands Execute these commands to validate the task is complete and works end 2 end: - Example: `uv run python -m py_compile apps/*.py`
File Write with User-Controlled Filenamemediumresource abuse
The skill writes the plan to `PLAN_OUTPUT_DIRECTORY/<filename>.md`, where the filename is generated based on the main topic of the plan (derived from user input). Insufficient sanitization of the filename could lead to overwriting critical files or creating files with unexpected extensions.
70% confidenceline 35
Save the complete implementation plan to `PLAN_OUTPUT_DIRECTORY/<descriptive-name>.md`
External System Integration for Social Engineering/Exfiltrationmediumsocial engineering
The plan format includes fields like 'External ticket subtask ID' and mentions creating subtasks in Jira, Linear, or GitHub. This implies potential interaction with external project management systems, which could be abused for social engineering (e.g., creating fake urgent tasks) or data exfiltration.
70% confidenceline 289
External ticket subtask ID <ID to the subtask in jira,linear or github for updating progress>
RAG Poisoning via Agent Definition Filesmediumrag poisoning
The agent identifies team members from `.claude/agents/team/*.md` files. If these agent definition files are compromised (e.g., via a supply chain attack), an attacker could poison the agent's knowledge base, leading to the deployment of malicious or altered sub-agents.
80% confidenceline 213
Identify from `.claude/agents/team/*.md` or use `general-purpose`.
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/Fleron/Claude-plugins/team-plan)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/Fleron/Claude-plugins/team-plan"><img src="https://mondoo.com/ai-agent-security/api/badge/github/Fleron/Claude-plugins/team-plan.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/Fleron/Claude-plugins/team-plan.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