principle-guard-the-context-window
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown content focused on architectural principles for token management. It does not contain any executable scripts, shell commands, or automation logic.- [NO_CODE]: No code or scripts are present in the skill files. The instructions are directed at the AI agent's reasoning strategy rather than system execution.- [SAFE]: The skill configuration uses
disable-model-invocation: true, which is a platform-level restriction that reduces the attack surface.- [SAFE]: The only external reference is a relative path to a local project file (../pstack-pi/references/runtime.md), which does not involve remote network access or untrusted source code execution.
Audit Metadata