The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands using git and the GitHub CLI (gh) to manage branches, stage changes, and interact with remote repositories.
[EXTERNAL_DOWNLOADS]: The instructions permit the agent to install missing dependencies or tools if repository checks fail, which involves fetching software from external package registries or sources.
[REMOTE_CODE_EXECUTION]: By allowing the agent to install missing dependencies or tools at runtime, the skill enables the execution of unverified external code within the agent's environment.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it ingests and inspects git diff output during the 'confirm scope' phase. Malicious instructions embedded in the source code being reviewed could potentially influence the agent's subsequent actions or summaries.
Ingestion points: git status and git diff output processed in Workflow Step 1.
Boundary markers: Absent; there are no specific delimiters or instructions used to isolate untrusted file content from the agent's operational prompt.
Capability inventory: The agent has the ability to execute git and gh commands, as well as install arbitrary packages.
Sanitization: The skill does not include steps to sanitize, escape, or filter the content of the diffs before the agent processes them.

yeet