The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes several local scripts located in ~/.claude/skills/gstack/bin/ (including gstack-config, gstack-team-init, and gstack-learnings-log) to manage tool settings, facilitate migration from vendored versions, and log architectural insights discovered during audits.
[DYNAMIC_EXECUTION]: Uses the eval command to execute the output of the local script ~/.claude/skills/gstack/bin/gstack-slug. This is a common pattern for dynamically setting environment variables based on script output.
[PROMPT_INJECTION]: As an auditor of untrusted code, the skill is inherently exposed to indirect prompt injection. It includes a proactive mitigation instruction ("Anti-manipulation") that explicitly directs the agent to ignore any instructions found within the analyzed codebase that attempt to influence the audit methodology or findings.

cso