openspec-archive-change

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface detected.
  • Ingestion points: The skill reads content from tasks.md and spec files within the openspec/ directory structure to assess completion and synchronization state.
  • Boundary markers: Absent. The instructions do not define delimiters or provide specific warnings to the agent to disregard potential instructions embedded within the processed file content.
  • Capability inventory: The skill utilizes shell command execution (via the openspec CLI, mkdir, and mv commands) and file system access.
  • Sanitization: None. The skill processes data from external files without validation or escaping logic.
  • [COMMAND_EXECUTION]: Potential for command injection through shell parameter interpolation.
  • The skill constructs shell commands using a variable name, such as openspec status --change "<name>" --json and mv openspec/changes/<name> ....
  • If the change name is derived from untrusted input and contains shell metacharacters (e.g., ;, &&, |), it could lead to the execution of arbitrary commands on the system.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 09:52 PM