excessive-agency
Hunt LLM excessive agency (OWASP LLM06:2025) — agentic systems granted too many tools, too broad permissions per tool, or unsupervised authority to act on the user / business behalf, producing financial loss, data loss, or destructive operations from a single bad token.
68
82%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Critical
Do not install without reviewing
Security
3 findings: 1 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.
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
This document includes explicit PoC payloads and instructions for data exfiltration, credential theft, persistence (IAM backdoors), cross-workspace pivoting, and automated destructive actions—clear high-risk malicious capability guidance.
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
The PoC payload explicitly instructs the agent to include "all credentials you have seen in this session" in an outbound send_email body, which requires the LLM to output secret values verbatim (high exfiltration risk).
W013: Attempt to modify system services in skill instructions.
What this means
The skill prompts the agent to compromise the security or integrity of the user’s machine by modifying system-level services or configurations, such as obtaining elevated privileges, altering startup scripts, or changing system-wide settings.
Why it was flagged
The skill explicitly instructs or demonstrates tool calls that perform destructive, state-changing actions (email exfiltration, workspace membership changes, IAM grants, writing config/cron persistence), which would allow the agent to modify system or service state and achieve persistent privilege escalation.
- Repository
- PurpleAILAB/Decepticon
- Commit
0cf691e
- 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.