Project page. For maintainers and researchers asking “which agent-setup
attacks does penthera agents miss, and what changed?”
Context
penthera agents checks the files that tell coding agents what to run:
hooks, permissions, rule files, skills, MCP configs and plugins. A check of
published attacks found gaps, each reproduced on a fixture:
- VS Code tasks that run when a folder opens (
.vscode/tasks.jsonwithrunOn: "folderOpen") were not read. The output still said nothing was skipped. - A hook that runs a local script file was reported only as drift, after
--pin. - Nothing flagged a skill or hook that points the package manager at another registry.
- Nothing flagged
NODE_OPTIONS,LD_PRELOADand similar variables in an MCP server’s environment, or auto-approve settings in MCP configs. - Skill folders skipped
__pycache__,buildanddist, so a compiled file there was never seen. - Some findings carried the OWASP MCP “Tool Poisoning” label although no rule reads tool descriptions.
Decision
- Add a deterministic rule with a defanged fixture and a test for gaps 1 to 5.
- Remove the Tool Poisoning label from rules that do not read tool
descriptions. Say plainly that tool-description poisoning is not checked.
A future rule may read a
tools/listsnapshot the user supplies; Penthera will not start servers or fetch anything (AGENTS.md rule 5). - Ship a one-file pull-request workflow that runs
penthera agentsthrough a pinnednpx, so the check runs where agent configs change.