The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from multiple external sources, creating an ingestion surface where instructions could be embedded in non-trusted data.
Ingestion Points: The skill ingests outputs from various subagents (lead-programmer, qa-tester, producer, qa-lead) and the results of external skill invocations (/smoke-check, /team-qa, /bug-report).
Boundary Markers: There are no explicit delimiters or instructions to ignore embedded commands within the data received from subagents or external tools.
Capability Inventory: The skill possesses significant capabilities, including Write and Edit for file modification, Bash for Git repository manipulation (merging, tagging), and the Agent tool to trigger further actions.
Sanitization: The instructions do not define any sanitization or validation logic for the content returned by subagents or testing tools before it is used to determine deployment readiness.
[COMMAND_EXECUTION]: The skill executes shell commands via the Bash tool for repository management and testing.
Evidence: It performs Git operations including git rev-parse, git tag, git branch, and git checkout to manage hotfix branches and releases.
Evidence: It executes testing commands dynamically retrieved from project.yaml (commands.test), which represents a potential vector if the configuration file is modified by an untrusted party.

hotfix