The Agent Skills Directory

[SAFE]: The skill outlines a standard developer workflow for debugging and fixing code failures. It utilizes well-known tools such as make, git, and pytest (via uv) for local operations. The instructions prioritize incremental fixes and verification, which aligns with development best practices. No evidence of obfuscation, unauthorized data access, or remote code execution was found.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it instructs the agent to analyze external data (test output).
Ingestion points: Test results and error messages from make test and pytest are analyzed by the agent (SKILL.md).
Boundary markers: Absent; the skill does not define specific markers to delimit test output from other instructions.
Capability inventory: The skill utilizes the Edit tool for file modifications and command execution via make and uv (SKILL.md).
Sanitization: Absent; the agent is expected to interpret raw terminal output. However, as this is the primary purpose of the skill and relies on the user's local environment, the risk is minimal and characteristic of the intended functionality.

test-fixing