$ loading_
Safely operate Directus data with schema-aware guards and verification controls.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "directus-safe-mcp" yet — see the docs or source repo.
Using directus-safe-mcp, run a dry-run batch update on the articles collection: change 20 records with status=draft to review. Return only the number of records that would change, sample records, and any safety-limit warnings. Do not commit changes.
A dry-run result showing the estimated changed count, sample affected records, and any batch-limit or safety-check warnings.
Use directus-safe-mcp to query the 10 most recent records from the customers collection, and state whether the collection is on the access allowlist. If it is not allowed, refuse clearly and explain why.
Either the query results or a clear refusal based on the collection allowlist, preventing unauthorized access.
Using directus-safe-mcp, update the price to 99 for id=123 in the products collection and run mutation verification. Output a before/after comparison of key fields and the verification result.
A verified mutation result showing whether the key field changes succeeded and passed verification.
Developers can use it to add schema awareness and safety guardrails when connecting Directus to an AI assistant. It is especially useful when collection access must be restricted and bulk mutations need tighter control.
Ops or platform teams can use dry-run, mutation verification, and batch limits to inspect changes before and after writes. This makes data updates more controlled and easier to audit.
Data analysts or business support users can rehearse changes before executing batch operations in Directus, reducing the chance of large-scale mistakes. Its batch caps and verification mechanisms help lower risk.
It is an MCP sidecar that wraps the Directus REST API with schema awareness and safety controls. It offers collection allowlists, dry-run support, batch limits, and mutation verification to make data operations more LLM-friendly.
Based on the description, it includes collection allowlists, batch limits, and dry-run plus verification guards for mutations. The core goal is to add constraints and checks around real data operations.
The provided material does not include installation steps or runtime prerequisites. It is clearly related to the Directus REST API; for exact setup details, see the source repository.
Secure LLM and MCP tool interactions with zero-trust controls and policy enforcement.
Safely inspect local dev databases with read-only schema, samples, and explain tools.
Let AI safely query and manage DuckDB databases through MCP tools.
Lets AI safely explore SQL schemas and run read-only database queries.
Safely explore PostgreSQL schemas and run read-only SELECT queries with AI.
Convert OpenAPI specs into secure MCP tools with guardrails and auditing.