mila
ReviewAudited by ClawScan on May 10, 2026.
Overview
This is a coherent instruction-only Mila integration, but it can use a Mila API key to read, change, or permanently delete cloud documents, sheets, and slides.
Install or use this skill only if you intend to let an agent manage Mila content. Use a dedicated least-privilege Mila API key, verify the api.mila.gg and mcp.mila.gg endpoints, keep MCP config files private, and require explicit confirmation before deletes or full-content replacements.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If an agent acts on the wrong item, it could permanently delete or replace important Mila content.
The skill documents permanent destructive operations through the API/MCP workflow. This is purpose-aligned CRUD behavior, but misuse or a wrong resource ID could cause irreversible data loss.
Deletes the workbook and all of its tabs. This is permanent.
Use explicit user confirmation for deletes and full replacements, verify resource IDs, and prefer least-privilege keys without write/delete scopes unless needed.
The agent can act within the permissions granted to the Mila key, potentially across personal files or team workspaces.
A Mila API key delegates account or team access to the agent. The artifact also describes read/write scopes, so the credential use is disclosed and purpose-aligned.
All requests require a [Mila](https://mila.gg) API key. Keys use the format `mila_sk_*`.
Create a dedicated Mila API key with only the scopes needed for the task, avoid broad team write/delete scopes when possible, and rotate or revoke the key if exposed.
A user relying only on registry metadata might not realize before opening the instructions that a cloud API key is needed.
The registry-level requirement fields do not declare the credential contract, while the skill instructions require a Mila API key. This appears to be an install-listing visibility gap rather than hidden behavior because SKILL.md discloses the API key requirement.
Required env vars: none ... Env var declarations: none ... Primary credential: none
Treat the skill as requiring network access and a Mila API key despite the registry requirement fields, and verify the endpoint and publisher before configuring credentials.
Mila document content, commands, and the API key may be sent through the configured Mila MCP connection.
The skill supports a remote MCP endpoint that receives an Authorization header. This is expected integration behavior, but it is a sensitive communication boundary.
MCP endpoint: `https://mcp.mila.gg` ... `"Authorization": "Bearer mila_sk_your_key_here"`
Use only the official HTTPS endpoint, keep the MCP config private, and use a scoped key that can be revoked.