The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration attempts were detected. The skill uses standard Elixir development tools (mix) and interacts with local project files (logs, tests, and solutions) as expected for a diagnostic tool.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from application logs and test outputs to identify root causes of bugs.
Ingestion points: Reads content from log/dev.log, mix test output, and external diagnostics via the Tidewave MCP.
Boundary markers: The skill does not explicitly use delimiters or "ignore embedded instructions" warnings when processing error messages or logs.
Capability inventory: The skill has the ability to execute shell commands (mix compile, mix test, mix ecto.migrate, grep) and spawn subagents.
Sanitization: There is no explicit sanitization of ingested error strings or log entries.
Context: This behavior is inherent to the skill's primary purpose as a debugging tool and is considered a normal operational surface.

investigate