The Agent Skills Directory

[SAFE]: The skill operates as a high-level analytical assistant for fiction writers. It focuses on identifying logic and continuity errors using established copyediting standards (e.g., Chicago Manual of Style). It explicitly states it does not modify the author's original prose.
[DATA_EXPOSURE]: The skill utilizes the Model Context Protocol (MCP) to access manuscript files. This access is consistent with its primary function (continuity checking across long texts). There are no network operations to non-whitelisted or suspicious domains.
[PROMPT_INJECTION]: The instructions follow a structured methodology for extracting and comparing facts. No instructions were found that attempt to bypass AI safety guardrails or override system instructions.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided text (manuscripts). While this creates an attack surface for indirect injection, the skill's capabilities are limited to reading, searching, and writing marginalia (non-executable feedback), and the instructions emphasize a 'query posture' where findings are presented for human review rather than acted upon autonomously.

continuity-editor