The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it reads and summarizes output from gh run view --log-failed, which can contain arbitrary text from CI logs controlled by the code being tested.
Ingestion points: External data enters the context via gh run list and gh run view output from the repository under test in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters to isolate the untrusted log data from the agent's system prompt.
Capability inventory: The skill utilizes Bash (for gh and git tools) and the Monitor tool to track run progress.
Sanitization: The skill does not describe any sanitization or validation logic for the content retrieved from the GitHub CLI.
[COMMAND_EXECUTION]: The skill uses bash variables to store and interpolate user-provided arguments into shell commands. Specifically, parameters like --branch, --sha, and --interval are passed directly into gh commands, which is a standard pattern that requires careful handling to prevent shell injection.

watch-ci