The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from multiple sources including $ARGUMENTS, Git logs, and external project files (plans, stories, issues). It lacks explicit boundary markers or sanitization logic to isolate these inputs from its internal instructions.
Ingestion points: $ARGUMENTS variable in SKILL.md; Git logs and project documents (plans, stories, issues) referenced in the Consolidation and Closure process steps.
Boundary markers: Absent; the skill does not instruct the agent to use delimiters or explicitly ignore instructions embedded within the processed content.
Capability inventory: The skill writes files to the local file system (planning/ directory) and executes shell commands for verification (lint, typecheck, tests).
Sanitization: Absent; no evidence of filtering or escaping logic for external data before it is incorporated into the prompt context.
[COMMAND_EXECUTION]: The Closure mode requires the agent to run verification commands including lint, typecheck, and tests. While these are legitimate developer workflows for an engineering-focused skill, they involve executing local shell commands whose behavior is determined by the project's own configuration files.

agile-status