The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown instructions, risk matrices, and YAML configuration. There are no scripts (Python, JavaScript, Shell) or binaries included in the skill package, which eliminates typical execution-based attack vectors.
[SAFE]: The skill is designed to enhance security and safety by enforcing 'human-in-the-loop' confirmations for destructive, irreversible, or high-impact operations. It includes comprehensive escalation rules for production environments and sensitive data access.
[INDIRECT_PROMPT_INJECTION]: While the skill processes context from the agent's current task (which could theoretically contain injected instructions), the skill itself lacks internal capabilities (such as network access or file writing) to be exploited. Its primary function is to act as a safety gate for other tools.

checkpoint-guardian