The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads data from local files (voice.md, swipe.md, log.md) to inform the content planning process. These files represent an ingestion surface where untrusted or maliciously crafted data could influence the agent's output.
Ingestion points: Reads local files located at ~/.claude/instagram/voice.md, swipe.md, and log.md.
Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the processed files.
Capability inventory: The skill is limited to reading files, prompting the user for input, and writing a text-based plan to ~/.claude/instagram/plan.md. It lacks network access or general command execution capabilities.
Sanitization: No evidence of sanitization or validation of the content read from the files before it is interpolated into the planning process.

ig-plan