The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends executing npx ruflo doctor --fix to address initialization errors, which triggers a download and execution from the NPM registry.
[COMMAND_EXECUTION]: Relies on the aqe command-line utility for core operations such as indexing (aqe kg index), searching (aqe kg search), and dependency mapping (aqe kg deps).
[DATA_EXFILTRATION]: The skill indexes the contents of the src/ directory, extracting entity relationships and generating embeddings that are stored in the AgentDB vector database.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it processes untrusted local source code for indexing and context generation.
Ingestion points: Processes all TypeScript files in the source directory (SKILL.md).
Boundary markers: No delimiters or instructions are provided to the agent to ignore embedded commands in the analyzed code.
Capability inventory: The skill uses the aqe tool to parse code and extract metadata, which is then used to generate intelligent context for AI operations (SKILL.md).
Sanitization: No sanitization or validation of the ingested code content is performed before it is used to construct knowledge graphs or search results.

qe-code-intelligence