The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard command-line utilities including git (for inspecting logs and diffs) and grep (for finding debug tags). These tools are used appropriately for root-cause analysis and do not exceed the necessary scope for the task.\n- [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface because it reads external data such as git commit history and error messages. While this data is untrusted, the skill's structured approach and the nature of the task (debugging) make this a routine operational risk rather than a security flaw.\n
Ingestion points: SKILL.md Step 1 involves reading output from git log and git diff.\n
Boundary markers: No specific boundary markers are defined to isolate external data from the agent's instructions.\n
Capability inventory: The skill uses shell commands for information gathering (git, grep) and suggests the use of local testing frameworks and curl for reproduction loops.\n
Sanitization: There is no specific instruction for the agent to sanitize or escape the content of the logs or diffs it reads.

diagnose