The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves 'context', 'operationGuidance', and 'rules' from external command outputs and instructs the agent to read and consider them when performing the archive and sync operations.
Ingestion points: The skill executes openspec instructions archive --json in Step 1 and openspec instructions specs --json in Step 4 to fetch data from the local repository.
Boundary markers: The skill contains explicit instructions to keep external fields separate from controlling inputs, report conflicts between guidance and command contracts, and treat guidance as advisory only.
Capability inventory: The skill has the capability to create directories (mkdir), move files and directories (mv), and trigger a secondary openspec-sync-specs workflow.
Sanitization: The instructions mandate that the agent should not copy rule text or guidance verbatim into outputs and must explain why if guidance is ignored due to conflicts.
[COMMAND_EXECUTION]: The skill uses standard shell commands to manage the file system during the archive process.
Evidence: Uses mkdir -p to create archive directories and mv to relocate the change root to the archive destination in Step 5.

openspec-archive-change