The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it incorporates untrusted external data into the agent's context.
Ingestion points: The skill fetches data from GitLab including Merge Request descriptions (glab mr view), file diffs (glab mr diff), issue details (glab issue view), and existing discussion threads (glab api).
Boundary markers: Absent. There are no instructions or delimiters provided to separate the agent's operational instructions from the external data being analyzed.
Capability inventory: The agent has the ability to execute shell commands and perform write operations (posting comments and discussions) back to the GitLab API.
Sanitization: Absent. The skill does not perform any validation or sanitization of the content retrieved from GitLab before processing it.
[COMMAND_EXECUTION]: The skill uses multiple shell commands via glab and git to perform its primary functions. While these are necessary for the skill's purpose, they provide a capability surface that could be exploited if the agent's instructions are subverted by malicious content in a Merge Request.

code-review