The Agent Skills Directory

[COMMAND_EXECUTION]: The helper script analyze_test_logs.mjs uses child_process.execSync to run Docker commands (docker compose ps, docker compose logs) to extract log data for analysis.
[EXTERNAL_DOWNLOADS]: The skill fetches reference files from the author's GitHub repository if they are not found in the local project directory. It also communicates with external Loki API endpoints to retrieve logs when configured.
[PROMPT_INJECTION]: The skill processes untrusted log data from external sources, presenting an indirect prompt injection surface.
Ingestion points: Log data gathered from Docker Compose, local filesystem globbing (*.log), and Loki query results.
Boundary markers: None explicitly defined in the skill's prompt construction, though the script parses and structures the data into JSON before it is processed by the AI.
Capability inventory: The skill can read local files, execute shell commands (specifically Docker), and perform network GET requests (fetch).
Sanitization: The script analyze_test_logs.mjs includes a normalization step that replaces dynamic content such as UUIDs, IP addresses, and timestamps with static placeholders to reduce noise and anonymize sensitive information.

ln-514-test-log-analyzer