The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration attempts were detected. The skill follows best practices by instructing the agent to avoid indexing secrets.
[COMMAND_EXECUTION]: The skill provides instructions for using a CLI tool (context-mode) to index directories. This is presented as a legitimate fallback mechanism when MCP tools are unavailable.
[DATA_EXPOSURE]: The skill is designed to process local file content into a persistent FTS5 knowledge base. It explicitly cautions against indexing sensitive data such as .env files, SSH keys, or other credentials, and recommends using exclusion paths to limit the scope of indexed data.
[DATA_EXPOSURE]: (Indirect Prompt Injection Surface Analysis)
Ingestion points: Local project files and directories (SKILL.md).
Boundary markers: Not explicitly defined in the instructions for the indexing process.
Capability inventory: Uses MCP tools (ctx_index, ctx_search) and shell command execution (context-mode index).
Sanitization: The skill relies on the user to exclude sensitive paths and does not specify internal sanitization of indexed content.

ctx-index