The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its ingestion of external and local data.
Ingestion points: The skill ingests data from external WebSearch results (Phases 1 and 2) and local codebase files scanned via grep (Phase 3).
Boundary markers: No delimiters or instructions to ignore embedded commands are implemented for the processed data.
Capability inventory: The skill has access to WebSearch and local file scanning tools.
Sanitization: No evidence of sanitization or validation of the retrieved content is present before it is used to generate hypotheses.
[COMMAND_EXECUTION]: The skill performs automated searches of the local environment using keyword patterns.
Evidence: The Phase 3 workflow and research_query_templates.md instruct the agent to use grep to find strings such as redis, memcache, pool_size, and max_connections in project source and configuration files.

ln-812-optimization-researcher