The Agent Skills Directory

[SAFE]: The skill is comprised of markdown instructions and workflow guidelines. Comprehensive analysis across all 9 files reveals no evidence of malicious patterns, obfuscation, or unauthorized network operations.
[DATA_EXPOSURE]: The instructions promote secure data handling. For example, in HANDOFF-PLANS.md, it explicitly mandates keeping secret values out of plans, and in PRODUCT-PASS.md, it warns against including secrets or PII in telemetry metrics.
[REMOTE_CODE_EXECUTION]: While the skill guides the agent through execution tasks (EXECUTION.md), it repeatedly emphasizes operating within the user's existing authorization and repository conventions, rather than downloading or executing untrusted external scripts.

engineering-planning