The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests data from local research files without sanitization.
Ingestion points: The skill reads research data from .ajtbd/segments.md and .ajtbd/job-graph.md using the Read tool.
Boundary markers: There are no specified boundary markers or delimiters to differentiate between the research data and the system instructions during processing.
Capability inventory: The skill is granted powerful capabilities including Read, Write, Glob, Bash, and WebSearch tools.
Sanitization: No sanitization, escaping, or validation of the ingested file content is performed, allowing potentially malicious instructions in the source files to influence agent behavior.

ajtbd-landing