The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the run_command tool to execute CLI utilities including git, gob-curl, bb, curl, and jq for repository and CI tasks.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch data from trusted developer infrastructure domains including dart-review.googlesource.com, cr-buildbucket.appspot.com, and logs.chromium.org.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and processing untrusted data (Gerrit review comments, commit messages, and build logs) without explicit boundary markers or sanitization.
Ingestion points: Data enters the agent's context through shell command outputs from gob-curl, bb, and curl targeting Gerrit and Buildbucket APIs.
Boundary markers: No explicit delimiters or isolation instructions are provided in the skill to separate the untrusted external data from the agent's primary instruction set.
Capability inventory: The agent has shell access via the run_command tool, which could be exploited if malicious instructions embedded in the data are followed.
Sanitization: The instructions do not provide methods for sanitizing or validating the fetched content before the agent processes it.

gob-curl