The Agent Skills Directory

[SAFE]: The skill provides instructional guidelines for the agent to follow a multi-step reasoning process. It does not include any scripts, binaries, or commands that interact with the host system or external networks.- [NO_CODE]: The provided files consist of markdown instructions and YAML configuration data only. There are no Python scripts, Node.js packages, or shell commands included in the skill.- [PROMPT_INJECTION]: While the skill contains structured instructions for agent behavior (e.g., 'Core Invariants' and 'Binding Rules'), these are design patterns for reasoning and do not attempt to override safety filters, bypass system constraints, or extract internal system prompts.- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for handling user-supplied tasks. Although this involves processing untrusted data, the skill lacks any capabilities (such as file writing, network access, or command execution) that could be exploited through injection. The 'Inner Register' and 'Ledger' act as internal state management rather than external-facing interfaces.- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or private secrets were found in the skill metadata or instructions.- [OBFUSCATION]: The text is clear and readable. No Base64, hex encoding, zero-width characters, or homoglyph substitutions were detected.

j-space