openspec-context-loading-cn
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by reading and summarizing external project documentation, which presents an attack surface for indirect prompt injection if the project files contain malicious instructions. \n
- Ingestion points: Files within the
spec/specs/andspec/changes/directories. \n - Boundary markers: None identified in the provided command templates. \n
- Capability inventory: File system access and text processing using standard utilities like
find,grep,cat,awk, andbc. \n - Sanitization: No content filtering or sanitization is applied to the ingested data. \n- [COMMAND_EXECUTION]: The skill utilizes shell commands to discover and display file contents. While these commands are standard and align with the skill's context-loading purpose, they involve filesystem interactions that rely on the agent to safely handle paths and identifiers.
Audit Metadata