The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its core function of reading and processing external codebase content.
Ingestion points: The agent reads the content of every file identified in the codebase during the Phase 3 'Systematic File Review'.
Boundary markers: Absent. There are no instructions or delimiters directing the agent to ignore potentially malicious instructions found within the audited files.
Capability inventory: The skill has the capability to execute shell commands (tsc, eslint, npm run lint) and write report files to the local filesystem.
Sanitization: None. The content is processed in its raw form for architectural and code quality analysis.
[COMMAND_EXECUTION]: The skill invokes several command-line utilities.
In Phase 2 and Phase 5, it executes tsc, eslint, and grep.
It triggers npm run lint, which executes scripts defined in the local environment's package.json file, potentially leading to the execution of unverified logic if the codebase itself is malicious.

audit