A local-first Model Context Protocol server that detects, reviews and redacts sensitive information before prompts reach AI assistants.
Hook
One of 6 in hooks.json
Runs when a session starts, running python -m securedact_enforced.gemini_hook --event SessionSta. From GigantesHJI/securedact-mcp.
Hook
One of 6 in hooks.json
Runs on the BeforeAgent event, running python -m securedact_enforced.gemini_hook --event BeforeAgen. From GigantesHJI/securedact-mcp.
Hook
One of 6 in hooks.json
Runs on the BeforeModel event, running python -m securedact_enforced.gemini_hook --event BeforeMode. From GigantesHJI/securedact-mcp.
Hook
One of 6 in hooks.json
Runs on the BeforeTool event for ^(mcp_.*, .*(http, web, search, fetch, request, api, connect).*, Read, Write, Edit, MultiEdit, NotebookEdit, Bash, Grep and Glob)$ tool calls, running python -m securedact_enforced.gemini_hook --event BeforeTool. From GigantesHJI/securedact-mcp.
Hook
One of 6 in hooks.json
Runs on the AfterTool event for ^(mcp_.*, .*(http, web, search, fetch, request, api, connect).*, Read, Write, Edit, MultiEdit, NotebookEdit, Bash, Grep and Glob)$ tool calls, running python -m securedact_enforced.gemini_hook --event AfterTool. From GigantesHJI/securedact-mcp.
Hook
One of 6 in hooks.json
Runs when a session ends, running python -m securedact_enforced.gemini_hook --event SessionEnd. From GigantesHJI/securedact-mcp.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: