The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill accesses sensitive configuration files, including .env and firestore.rules, as part of its cross-layer consistency audit. It also reads and writes to project-specific telemetry files located in ~/.claude/projects/.
[COMMAND_EXECUTION]: The skill dynamically constructs and executes numerous shell commands based on the detected project stack. This includes running package managers and build tools like npx, cargo, go, gradlew, and mvn, as well as static analyzers and testing frameworks.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes arbitrary codebase content and is instructed to 'fix' identified issues.
Ingestion points: The entire project directory provided in $ARGUMENTS is analyzed by the agent.
Boundary markers: No explicit delimiters or instructions to ignore instructions embedded in the code are used when the agent processes external files.
Capability inventory: The skill has significant capabilities, including automated file modification, git operations, and execution of complex build/test command sequences.
Sanitization: The skill does not implement sanitization or validation of the codebase content before it is incorporated into the agent's context or used to drive 'self-healing' actions.

analyze