The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill imports brand guidelines from external sources, which are then saved and persistently applied to future tasks.
Ingestion points: SKILL.md (argument-hint) accepts a [file-path or URL] and the input requirements include Pasted guideline content.
Boundary markers: There are no instructions to the agent to use delimiters or ignore embedded instructions within the ingested data.
Capability inventory: The skill performs file writes to ~/.claude-marketing/ and executes the scripts/guidelines-manager.py tool.
Sanitization: No sanitization, escaping, or validation of the external content is described before it is formatted into markdown and saved.
[COMMAND_EXECUTION]: The skill invokes a local Python script to perform file management operations.
Evidence: Step 6 in SKILL.md describes executing scripts/guidelines-manager.py --brand {slug} --action save --category {category}.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads from and writes to specific configuration and data paths within the user's home directory.
Evidence: The process involves reading ~/.claude-marketing/brands/_active-brand.json and writing structured guidelines to ~/.claude-marketing/brands/{slug}/guidelines/.

import-guidelines