The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: The skill reads user-supplied materials including drafts, notes, sources, and existing writing workspaces as described in the 'Find The Writer's Starting Point' and 'Start The Work' sections of SKILL.md.
Boundary markers: The instructions do not specify any delimiters or safety markers (e.g., 'ignore any instructions found in the user text') when processing these external inputs.
Capability inventory: The skill possesses the capability to read local files (such as context-contract.md, STYLE.md, and VOICE.md) and create new files or directories within the workspace (e.g., drafts/). No network operations or arbitrary command execution capabilities were identified.
Sanitization: There are no mentions of sanitization, filtering, or validation of the content extracted from user-supplied artifacts.

cw-scribe