The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands for file discovery and metadata extraction, including find, grep, git status, and ls.
[COMMAND_EXECUTION]: External developer tools pydeps and madge are invoked to perform dependency graph analysis and circular dependency detection.
[PROMPT_INJECTION]: The skill has a significant attack surface for Indirect Prompt Injection. \n
Ingestion points: The workflow involves reading and analyzing Architecture Decision Records (ADRs) and source code across the repository. \n
Boundary markers: No delimiters or safety instructions are used to distinguish analyzed code content from the agent's internal instructions. \n
Capability inventory: The agent can execute shell commands, run python/node analysis tools, and access the local file system. \n
Sanitization: External data from the analyzed codebase is processed without explicit sanitization or filtering.
[COMMAND_EXECUTION]: The coupling-analysis.md module utilizes python -c to execute an inline script that processes data piped from grep.
[COMMAND_EXECUTION]: The skill includes instructions to execute pytest on its own test suite for logic verification.

architecture-review