The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted ReactLynx source code provided by users.
Ingestion points: The skill instructions in SKILL.md direct the agent to use rg (ripgrep) and a Node.js-based scanner to read file contents from the repository into the agent's context.
Boundary markers: The instructions do not specify any delimiters or directives to prevent the agent from following instructions embedded within the source code being analyzed.
Capability inventory: The agent can execute shell commands (node, rg) and perform file system modifications during the refactoring process.
Sanitization: The input code is parsed by a local script using regex, but the agent reads the raw file content without sanitization.
[DYNAMIC_EXECUTION]: The SKILL.md file instructs the agent to execute a bundled JavaScript scanner (scripts/index.mjs) using the node -e command to perform code analysis.
[COMMAND_EXECUTION]: The workflow relies on shell commands such as rg for file searching and node for running the logic in the skill's utility scripts.

reactlynx-best-practices