The Agent Skills Directory

Command Execution: The skill performs shell operations using npx, mkdir, and mv to manage the archival process. While these are functional requirements for moving directories and checking tool status, they involve direct interaction with the host environment.
Potential for Indirect Prompt Injection: The skill reads external content from tasks.md and spec files within the openspec/changes/ directory to facilitate conflict resolution. If these user-controlled files contain instructions intended to influence the agent, they could potentially affect how the agent interprets implementations or resolves spec conflicts.
Ingestion points: openspec/changes/<name>/tasks.md and capablity spec files in openspec/changes/<name>/specs/.
Boundary markers: The instructions do not specify the use of boundary markers or delimiters when reading these files.
Capability inventory: The skill possesses capabilities for shell command execution (mkdir, mv, npx) and codebase searching.
Sanitization: There is no explicit requirement for sanitizing the input from these project files before the agent processes them.
External Package Usage: The skill utilizes npx openspec, which may involve downloading the openspec package if it is not already present in the environment. This is a standard pattern for CLI-based agent skills.

openspec-bulk-archive-change