The Agent Skills Directory

[COMMAND_EXECUTION]: The skill triggers shell-based git operations including git diff, git log, and git rev-parse based on user-supplied branch or commit references. It depends on the environment's tool execution safety. Evidence found in SKILL.md under the 'Process' section.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external content (code diffs and specification files) which may contain adversarial instructions targeting the agent's review logic.
Ingestion points: Git diff output and various repository files (docs/, specs/) are used as primary input for review sub-agents.
Boundary markers: Absent. Contents are interpolated directly into prompts without protective delimiters or instructions to disregard embedded commands.
Capability inventory: The skill possesses the capability to execute shell commands and read local file system data.
Sanitization: The skill lacks mechanisms to sanitize or validate the content of processed diffs and documentation before it is analyzed by sub-agents.

code-review