Vulnerability · Published Sep 4, 2026 · Assigned by VulnCheck
CVE-2026-85685AgentScope through 2.0.7.post1 Arbitrary Directory Copy via add_skill
High 7.5
In plain English
A high-severity path traversal in agentscope from agentscope-ai. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 0.5% (higher than 46% of all known vulnerabilities).
Official description
AgentScope through 2.0.7.post1 contains a path traversal vulnerability in LocalWorkspace.add_skill that copies arbitrary server directories into the agent workspace via an unconfined source path parameter. Attackers can supply any directory path in the skill_path request parameter to copy files into the skills directory, making them accessible through the workspace skill listing.
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| agentscope-ai | agentscope | 0 – through 2.0.7.post1 |
Weakness: CWE-22 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
References
- GitHub Issue #2069 issue-tracking
- https://github.com/agentscope-ai/agentscope product
- https://github.com/agentscope-ai/agentscope/blob/v2.0.7.post1/src/agentscope/workspace/_local_workspace.py technical-description
- VulnCheck Advisory: AgentScope through 2.0.7.post1 Arbitrary Directory Copy via add_skill third-party-advisory