The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like ls, cat, and grep to manage files and track project state. Examples include verifying project specifications with ls -la project-specs/*-setup.md and counting tasks using grep -c in the SKILL.md file.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and act upon data from external files (e.g., project-specs/[project]-setup.md). Malicious instructions embedded in these files could potentially influence the orchestrator's behavior or the instructions passed to spawned agents.
Ingestion points: Files located in the project-specs/ and project-tasks/ directories (referenced in SKILL.md).
Boundary markers: Absent. The skill does not define specific delimiters or instructions to ignore potentially malicious content within the specifications.
Capability inventory: The agent can execute shell commands (ls, cat, grep) and has the capability to spawn and provide instructions to multiple specialized agents (e.g., Frontend Developer, DevOps Automator).
Sanitization: Absent. There is no evidence of content filtering or validation for the data ingested from specification files.

agency-agents-orchestrator