The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local project source files (React components, CSS, parent components) and user-supplied bug descriptions during the diagnosis phase. Malicious instructions or hidden prompts embedded in these files could potentially influence the agent's reasoning or the resulting fix plan.
Ingestion points: Bug descriptions, component source files, story files, and related styling/parent files processed in SKILL.md (Steps 1, 2, and 4).
Boundary markers: The skill does not employ explicit delimiters or system instructions to disregard embedded commands within the ingested source code.
Capability inventory: The skill can execute shell commands via Bash, write files to $JAAN_OUTPUTS_DIR, and trigger subsequent skills through the Task tool (e.g., dev-output-integrate).
Sanitization: There is no evidence of filtering or sanitization of external content before it is interpolated into the agent's diagnostic context.
[COMMAND_EXECUTION]: The skill utilizes source to load internal utility scripts (id-generator.sh, index-updater.sh) located within the ${CLAUDE_PLUGIN_ROOT}. It also performs standard directory operations like mkdir and stat via the Bash tool. These operations are restricted to the plugin's own library paths and the designated output directory, representing standard internal functionality.

frontend-component-fix