The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill consists entirely of instructional text for maintaining storytelling consistency.
[PROMPT_INJECTION]: The instructions do not contain attempts to bypass safety filters, override system prompts, or use common jailbreak techniques.
[DATA_EXFILTRATION]: There are no network operations (curl, wget, fetch) or access to sensitive local file paths (e.g., credentials, SSH keys) observed.
[OBFUSCATION]: No evidence of Base64 encoding, zero-width characters, homoglyphs, or hidden text patterns was found.
[REMOTE_CODE_EXECUTION]: The skill does not perform package installations or download and execute external scripts.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process narrative data and configuration files like knowledge/LEDGER.md.
Ingestion points: Narrative content (scenes, chapters) and knowledge/LEDGER.md (specified in SKILL.md).
Boundary markers: None explicitly defined in the instructions.
Capability inventory: Uses narrative analysis tools such as timeline, mystery-engine, narrative-graph, power-system, and canon-manager. No file-writing or network capabilities are exposed by this specific skill.
Sanitization: None specified. The skill is low risk as it is focused on logical consistency analysis within the agent's context.

continuity