The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Node.js scripts, specifically .agents/scripts/resolve_agents.js and .agents/scripts/match_methods.js, to perform its core functions. These scripts are not included in the provided files, making their logic unverifiable.
[COMMAND_EXECUTION]: A command injection vulnerability is present in Step 1. The instruction node .agents/scripts/match_methods.js --context "[detailed task context, artifact name, and current section]" directs the agent to insert raw content into a shell command. If an attacker-controlled document contains shell metacharacters like semicolons or backticks, it could lead to arbitrary command execution on the host system.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. It ingests content from methods.csv (the description and output_pattern fields) and user-provided document sections but lacks boundary markers or sanitization instructions to prevent the agent from obeying malicious commands embedded in that data.

elicitation