The Agent Skills Directory

[SAFE]: The skill uses local file-system tools (Read, Glob, Grep, Write, Edit) to perform its intended functions of auditing and reporting on test coverage. No malicious patterns such as credential exfiltration, external downloads, or privilege escalation were detected.
[PROMPT_INJECTION]: A potential surface for indirect prompt injection exists because the skill processes content from bug reports and design documentation which could be influenced by external actors.
Ingestion points: design/gdd/systems-index.md, design/gdd/*.md, and production/qa/bugs/*.md in SKILL.md.
Boundary markers: Absent; the agent is not instructed to ignore commands embedded within the data files.
Capability inventory: The agent possesses Write and Edit capabilities used to update tests/regression-suite.md.
Sanitization: No sanitization of ingested text is performed before it is included in the output manifest.

regression-suite