The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill is specifically designed to ingest and process untrusted external strings (input_data). This presents a surface for indirect prompt injection where malicious instructions embedded in the data could influence the agent's behavior during the sanitization process.
Ingestion points: input_data field defined in skill.yaml.
Boundary markers: No specific delimiters or "ignore instructions" warnings are defined in the schema to isolate the untrusted data from the agent's operational logic.
Capability inventory: Based on the provided files, the skill lacks side-effect capabilities such as file writing, network access, or command execution, which limits the potential impact of an injection to the agent's internal reasoning or output generation.
Sanitization: No logic is provided to sanitize or escape instructions within the processed text; it focuses solely on PII redaction.
[No Code Execution] (INFO): The skill consists of a YAML configuration and a Markdown documentation file. No implementation logic (Python, Node.js, or Shell) was provided. The analysis is limited to the defined metadata and intended behavior.

pii-sanitizer