The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements a monitoring phase that ingests git logs and test outputs to assess loop progress. This creates a surface for indirect prompt injection where adversarial content in the repository history or test errors could influence the agent's orchestration advice.\n
Ingestion points: Phase 3 instructions involve running git log --oneline and observing test results to detect stuck loops.\n
Boundary markers: Absent; the instructions do not provide delimiters or safety warnings for the agent when processing these external data sources.\n
Capability inventory: The skill uses shell commands for monitoring and provides instructions for loop control and prompt adjustment.\n
Sanitization: Absent; there is no mention of validating or escaping the content retrieved from git logs or test outputs.\n- [COMMAND_EXECUTION]: Provides instructions for the agent to execute git log and presents a command-line template for the user to launch the ralph-loop tool. These are routine development activities within the context of the skill's stated purpose.

ralph-orchestrator