The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several shell commands to collect context:
gh issue view is used to retrieve structured data about a specific GitHub issue.
ls ai-workspace/decisions/*.md and tail -20 ai-workspace/MEMORY.md are used to read local project configuration and history files.
[EXTERNAL_DOWNLOADS]: The skill fetches issue data (title, body, labels) from GitHub. This communication is performed via the official GitHub CLI (gh) targeting a well-known service for the intended purpose of task management.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes external, untrusted content from GitHub issues without explicit sanitization or boundary markers.
Ingestion points: Data is ingested from the output of gh issue view (external issue title and body) and local workspace files like .branch-context.md and MEMORY.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when the agent processes the fetched issue context.
Capability inventory: The skill has the capability to execute shell commands (gh, ls, tail, head).
Sanitization: There is no evidence of sanitization, filtering, or validation performed on the issue content before it is used to determine task types or branch names.

orient