The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes data from untrusted external sources.
Ingestion points: Web discovery results (Phase 3.1), LinkedIn profiles, Glassdoor reviews, and news articles (Phase 3.2).
Boundary markers: Absent. The skill does not instruct the agent to ignore instructions embedded in the external content.
Capability inventory: Local file reading (config and service definitions), web search, and local file writing (markdown reports).
Sanitization: Absent. Data from the web is parsed and directly incorporated into the scoring and reporting process.
[DATA_EXFILTRATION]: The skill transmits business criteria (industry names, pain points, service categories) to external search engines (Phase 2.1). This is a functional requirement of the discovery process and does not involve the transmission of sensitive user credentials, private files, or system environment data.
[COMMAND_EXECUTION]: The skill performs file system operations including reading from .jobops/config.json and writing reports to paths specified within that configuration. These operations are limited to the skill's defined workspace and follow standard patterns for tool-specific configuration management.

findclient