The Agent Skills Directory

[NO_CODE] (SAFE): The skill consists entirely of Markdown instructions and metadata. It does not include any scripts, binaries, or configuration files.
[PROMPT_INJECTION] (SAFE): No instructions designed to override agent behavior, bypass safety filters, or extract system prompts were detected.
[DATA_EXFILTRATION] (SAFE): No hardcoded credentials (API keys, tokens) or attempts to access sensitive local file paths (e.g., SSH keys, env files) were found.
[REMOTE_CODE_EXECUTION] (SAFE): There are no patterns involving the download or execution of remote scripts or the installation of unverified packages.
[COMMAND_EXECUTION] (SAFE): The skill does not define any shell commands, subprocess calls, or system-level operations.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill is designed to process untrusted data from external sources like GitHub issues and technical blogs. While this creates a vulnerability surface, the risk is inherent to the skill's purpose and is mitigated by the absence of executable capabilities.
Ingestion points: External documentation URLs, GitHub issues, PRs, and technical blogs.
Boundary markers: Uses structured Markdown templates (Type A, B, C) to separate source content from agent analysis.
Capability inventory: None. The skill has no file-write, network-send (other than search), or code execution capabilities.
Sanitization: Not explicitly defined in the prompt instructions.

librarian