The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data such as PR diffs and doc files provided through the $ARGUMENTS parameter. This creates a surface for indirect prompt injection where malicious instructions embedded in source files could influence the agent's behavior during the sync process.
Ingestion points: Doc file paths and PR diffs provided as input to the skill.
Boundary markers: Absent. The skill does not provide instructions to the agent to disregard instructions or control sequences found within the source material.
Capability inventory: The skill performs file system reads and writes, and executes shell commands (bun lint:fix).
Sanitization: Absent. Source data is incorporated into skill files without filtering or escaping of potential prompt directives.
[COMMAND_EXECUTION]: The skill triggers the execution of external shell commands as part of its verification workflow.
Evidence: The verification section in SKILL.md explicitly calls for running bun lint:fix after modifications are applied.

sync-skill