The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of local shell commands via the hunk CLI. This includes inspecting session state (hunk session list, get), navigating file diffs (hunk session navigate), and modifying session contents (hunk session reload). It also uses pipes to send JSON payloads to the CLI (printf ... | hunk session comment apply).
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by ingesting untrusted data from the local repository and live diff sessions through hunk session review --json and hunk session context.
Ingestion points: Diff content and file structures are read into the agent's context in SKILL.md via the review and context commands.
Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions embedded within the code diffs being reviewed.
Capability inventory: The agent can execute commands (hunk session reload) and write data back to the session (hunk session comment add/apply).
Sanitization: There is no mention of sanitizing or escaping the diff content before processing.

hunk-review