The Agent Skills Directory

[PROMPT_INJECTION]: The skill has an attack surface for Indirect Prompt Injection.
Ingestion points: Untrusted data enters the context via notion_notion-fetch used in SKILL.md to retrieve page content.
Boundary markers: Absent. The fetched content is not delimited or marked as untrusted for the agent.
Capability inventory: The skill possesses write capabilities to Notion (notion_notion-update-page) and shell command execution (cat).
Sanitization: Absent. There is no evidence of content validation or escaping for the data retrieved from Notion.
[DATA_EXFILTRATION]: The skill contains hardcoded identifiers for sensitive resources.
The .skill.config file includes specific Notion Page IDs for sensitive categories like Admin/Legal and Investor info.
The configuration includes links to well-known services (Firstbase and Cal.com) that contain specific company and team account identifiers.
[COMMAND_EXECUTION]: Shell commands are used for managing skill configuration.
The skill uses cat to read and write environment variable files at .opencode/skill/company-admin/.env.

company-admin