The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied material such as session notes, recordings, or chat logs and interpolates this content into structured Markdown output. Malicious instructions embedded in these inputs could potentially influence the agent's behavior or output during the summarization process.
Ingestion points: The skill explicitly requests raw material from the user (notes, recordings, chat logs) as the primary input for its processing logic.
Boundary markers: While the skill uses Markdown headers (e.g., ## Recap, ## NPCs) to structure the resulting note, it lacks explicit instructions for the agent to ignore or delimit embedded natural language instructions within the source data.
Capability inventory: The skill is designed to format text and can optionally interface with external notes applications or the 'Scratchpad connector' if available to save the resulting summary.
Sanitization: No evidence of sanitization, escaping, or validation of the input data is present in the instructions.
[NO_CODE]: The skill consists entirely of instructional text and Markdown templates. It does not include or execute any external scripts, binaries, or code-based logic.

campaign-notes