The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface due to its core data ingestion workflow.\n
Ingestion points: Requirements are ingested from Jira, GitHub, and arbitrary URLs via the WebFetch tool as described in SKILL.md (Step 0.1) and references/spec-workflow.md.\n
Boundary markers: The skill lacks explicit delimiters or isolation instructions when interpolating raw external requirements into the subagent prompts defined in references/agent-catalog.md.\n
Capability inventory: The agent team has access to powerful tools including Bash, Write, Edit, and the ability to spawn further subagents via TaskCreate and Teammate operations.\n
Sanitization: No sanitization or content filtering is performed on ingested external data before it is processed by the AI agents.\n
Mitigation: The workflow includes a mandatory USER APPROVAL GATE in Phase 5, requiring the user to review the full technical plan and task breakdown before any code changes are implemented in Phase 6.\n- [COMMAND_EXECUTION]: The skill utilizes shell commands and system CLI tools as part of its primary development workflow.\n
The orchestrator executes gh and jira CLI tools to ingest external issue data.\n
Implementation subagents (Frontend and Backend Developers) utilize the Bash tool to execute migrations, run tests, and perform codebase discovery as part of the intended development team functionality.

team-implement