The Agent Skills Directory

[SAFE]: The skill's primary function is to perform a deterministic scan of the local filesystem. It uses standard Node.js APIs to identify project types and tooling without external dependencies or network access.
[DATA_EXFILTRATION]: The script reads wp-config.php to identify if specific WordPress constants (e.g., WP_DEBUG) are enabled. While this file is highly sensitive, the analysis is restricted to identifying specific boolean flags and does not extract or report sensitive credentials like database passwords or keys.
[PROMPT_INJECTION]: The tool ingests content from external files (e.g., plugin and theme headers) that may be attacker-controlled. 1. Ingestion points: scripts/detect_wp_project.mjs reads local .php, .css, and .json files. 2. Boundary markers: The output is produced as structured JSON, but the script does not include explicit delimiters or warnings to the agent to ignore instructions within the extracted text fields. 3. Capability inventory: The script is limited to read-only filesystem operations and lacks network access, subprocess execution, or dynamic code evaluation capabilities. 4. Sanitization: Output is serialized using JSON.stringify, which prevents structural breaking of the JSON report but does not sanitize the semantic content of the extracted strings.

wp-project-triage