The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to read and analyze untrusted data (source code and diffs) from the repository. This creates a surface for indirect prompt injection where malicious instructions embedded in code comments or strings could influence the agent's review or trigger actions via available tools.\n
Ingestion points: The skill uses git diff and the Read tool to pull content from the current working directory and branch history into the agent's context.\n
Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings for the code being read.\n
Capability inventory: The skill has access to Bash, Write, Task, and Agent (for subagent delegation) tools.\n
Sanitization: There is no instruction to sanitize or filter the code content before it is processed by the agent or subagents.\n- [COMMAND_EXECUTION]: The skill uses the Bash tool to perform git operations (such as git status, git diff, and git symbolic-ref) and explicitly suggests running the repository's test command to verify code correctness. This results in the execution of scripts or binaries defined within the repository files being reviewed.

reviewkit