MakeX
WarnAudited by ClawScan on May 10, 2026.
Overview
This skill is coherent, but it needs review because it can use an organization token to run broad third-party actions without documented scoping or confirmation safeguards.
Use this skill only if you trust MakeX/Composio and need broad organization-level integration automation. Before using it, restrict connected integrations and action types, protect the X-Org-Token carefully, require human confirmation for mutating actions, and avoid using the output-structure endpoint on actions that can change real data.
Findings (5)
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.
An agent with the organization token and account ID could run available connected-service actions, such as sending messages or modifying third-party account data, depending on what integrations are connected.
This documents a general action-execution endpoint for connected third-party accounts, but the artifacts do not describe action allowlists, permission limits, user confirmation, or dry-run controls.
POST /api/openclaw/integrations/run-action ... Execute a specific action on a connected account ... toolName (required): The action slug to execute ... custom_auth_params (optional) ... custom_connection_data (optional).
Require explicit user approval for mutating actions, restrict allowed integrations and action slugs, prefer least-privilege tokens, add dry-run/read-only modes where possible, and log/audit all executions.
A user or agent trying to inspect an action's response shape could accidentally perform the action, such as sending a test email or changing connected-service data.
The output-structure endpoint sounds like an inspection helper but actually executes the action, which can create side effects if used with mutating actions.
Purpose: Execute an action and return its output structure. Useful for determining the shape of an action's response data.
Use metadata-only schema endpoints for inspection, or require a clear confirmation and a safe test account before any output-structure call that executes a real action.
If the token is exposed or overused by an agent, it could grant broad access to connected organizational integrations rather than a single narrow task.
The skill relies on an organization-level service token, which can authorize actions against organization-scoped connected accounts; the artifacts do not describe fine-grained scopes or safeguards.
All endpoints require the X-Org-Token header containing an organization service token ... Copy your API key from the settings page — this is your X-Org-Token.
Use least-privilege organization tokens, rotate tokens regularly, avoid placing tokens in prompts or logs, and ensure the skill metadata clearly declares the credential requirement.
Data supplied to actions and returned from connected services may pass through MakeX/Composio and could include sensitive organization information.
Forwarding action arguments and returning raw provider responses is expected for this integration, but those payloads may contain sensitive message, account, or repository data.
Calls Composio API: POST https://backend.composio.dev/api/v3/tools/execute/{toolName} ... Returns the raw Composio response.Send only the minimum necessary data, avoid optional custom authentication parameters unless required, keep tracing disabled unless needed, and review provider privacy/logging settings.
Users must trust the remote service behavior because the skill package itself does not include implementation code for independent review.
There is no local code package to inspect, and the remote MakeX/Composio implementation is not verifiable from the supplied artifacts.
Source: unknown; Homepage: none; No code files present — this is an instruction-only skill. The regex-based scanner had nothing to analyze.
Install only if you trust the publisher and MakeX service, and prefer skills that provide clear provenance, documentation, and source links for high-impact integrations.
