The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands for investigating software defects. This includes performing git archeology (using git log, git blame, and git diff), checking project dependencies (using npm ls or pip3 show), and running arbitrary reproduction tests in the terminal. These actions are within the expected scope of a software investigation tool.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external sources.\n
Ingestion points: The agent gathers evidence from error messages, stack traces, test outputs, and log entries during the 'Characterize' step (SKILL.md).\n
Boundary markers: The instructions do not define clear delimiters or use 'ignore' warnings for the external data processed during the investigation.\n
Capability inventory: The skill has access to powerful capabilities, including shell command execution via the Bash tool and the ability to spawn subagents using different models (SKILL.md).\n
Sanitization: There are no instructions for sanitizing or escaping the content of logs or error messages before they are processed by the agent or subagents.

investigate