The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The monorepo analysis tool ingests untrusted configuration data from the local filesystem to map dependencies and identify project types.
Ingestion points: The scripts/monorepo_analyzer.py script reads and parses package.json and pnpm-workspace.yaml files from the target directory.
Boundary markers: The script output lacks explicit boundary markers or instructions to the agent to disregard embedded content within the analyzed files.
Capability inventory: The skill is designed to interact with the filesystem and execute build/test commands via pnpm, turbo, and nx.
Sanitization: While the script uses standard JSON and YAML parsers, it does not explicitly sanitize the content before using it to define the dependency graph logic.
[DYNAMIC_EXECUTION]: The documentation includes instructions for the agent to generate and then immediately execute a Node.js script to create visualization diagrams.
Evidence: In references/monorepo-tooling-reference.md, a shell command sequence creates scripts/gen-dep-graph.js via a heredoc and subsequently executes it using node. This follows a simple script generation pattern from a known template for utility purposes.

monorepo-navigator