The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from external sources and possesses capabilities to modify the environment.
Ingestion points: External GitHub issue content is retrieved using gh issue view {{ARG1}} in SKILL.md.
Boundary markers: The instructions do not define clear delimiters or specific instructions for the agent to ignore potentially malicious content embedded within the issue description.
Capability inventory: The skill has the ability to create new GitHub issues via gh issue create and execute a local shell script (add-sub-issue-to-issue.sh) to modify issue relationships.
Sanitization: There is no explicit evidence of sanitization, filtering, or validation of the ingested issue content before it is processed or used to generate sub-issues.
[COMMAND_EXECUTION]: The skill executes local commands and a bundled script to perform its primary function.
Evidence: In SKILL.md, the skill uses gh issue view, gh repo view, and gh issue create to interact with GitHub.
Evidence: It also executes a script located at ${CLAUDE_PLUGIN_ROOT}/bin/add-sub-issue-to-issue.sh to link sub-issues. This is an expected behavior for the skill's intended purpose of managing GitHub issue hierarchies.

plan-task