The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted data from Jira issues and local logs, creating a surface for instructions embedded in that data to influence agent behavior.
Ingestion points: The workflow in SKILL.md (Phase 1) and references/discovery-intake.md instructs the agent to read User Story descriptions, Acceptance Criteria (AC), and Jira comments to derive test scenarios.
Boundary markers: There are no explicit prompt delimiters or instructions defined to isolate the ingested Jira content from the agent's internal control instructions.
Capability inventory: The skill possesses extensive write capabilities in the target TMS environment, utilizing [ISSUE_TRACKER_TOOL] and [TMS_TOOL] to create, link, and transition Jira/Xray issues.
Sanitization: As a primary security control, the skill implements a mandatory human "signature" requirement (described in SKILL.md and references/tms-conventions.md) before any automation verdict is finalized or promoted, effectively placing a manual review gate on the most critical actions.

test-documentation