The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and act upon data from untrusted external sources.
Ingestion points: The skill instructions in SKILL.md explicitly direct the agent to query live issue/project systems, pull requests, commits, and external branches to build a dependency graph and dispatch tasks.
Boundary markers: While the skill includes an 'EXCLUDED' mechanism to protect specific files or items, it lacks explicit delimiters or instructions to ignore potential malicious prompts embedded within external issue descriptions or code comments.
Capability inventory: The skill possesses high-impact capabilities including file system writes, git workflow management (commit, merge, push), and authorized deployment/publication actions.
Sanitization: There is no evidence of sanitization or validation of external task data before it is interpolated into worker prompts.

gogopowermanager