The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted data such as logs and stack traces, which creates a surface for indirect prompt injection.
Ingestion points: Technical data provided by the user is stored in artifacts/ and report.md for processing by sub-agents in Phase 2.
Boundary markers: The skill does not define specific delimiters or 'ignore' instructions for the sub-agents analyzing these external logs.
Capability inventory: The agent possesses capabilities to read local source code, orchestrate sub-agent workflows, and perform write operations (issue creation/editing) via the gh command.
Sanitization: No explicit sanitization or filtering logic is provided for the contents of the logs or stack traces prior to AI analysis.
[COMMAND_EXECUTION]: The skill utilizes system commands to perform its core functions, including git log, git blame, and the gh (GitHub) CLI.
Execution Scope: Commands are limited to read-only operations for code analysis and specific issue-management tasks (gh issue create, gh issue edit, gh label create).
Safety Protocols: A mandatory human-in-the-loop approval step (Phase 7) is required before any write operations are executed on GitHub. If the gh command is unavailable, the skill falls back to providing text for manual submission.

issue-report-dev