aci-mcp
Provides two ways to run MCP (Model Context Protocol) servers that expose ACI.dev managed functions/tools to MCP clients: an apps-specific server that exposes tools from selected apps, and a unified server that can discover and execute tools across all available functions (via meta-tools).
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Security posture is partially supported by using an API key (via environment variable in examples). However, the provided documentation does not describe transport guarantees (TLS), token scopes/least-privilege, rate limiting, or how sensitive errors are handled. Unified-server exposure/discovery may increase the blast radius if used with overly broad access.
⚡ Reliability
Best When
You want an MCP-compliant integration to ACI.dev tools and can supply an ACI API key plus linked account owner ID, optionally restricting apps via CLI flags.
Avoid When
You need strict least-privilege at the tool/function level beyond what the apps-server restriction provides, or you cannot run a local MCP server process/container.
Use Cases
- • Integrate an AI agent with ACI.dev tools via the MCP protocol
- • Expose a curated subset of ACI.dev app tools to an MCP client (apps-server)
- • Enable discovery + execution of ACI.dev tools through a unified MCP server (unified-server)
Not For
- • High-sensitivity environments without reviewing tool exposure/discovery behavior
- • Use cases requiring REST/GraphQL/gRPC APIs or webhooks from this package itself
Interface
Authentication
Authentication appears to be handled via an ACI API key (ACI_API_KEY). Exact scope granularity and token format are not specified in the provided README snippet.
Pricing
No pricing information found in provided README/metadata.
Agent Metadata
Known Gotchas
- ⚠ Unified server claims 'unlimited tool access' and includes tool discovery/execution meta-functions; an agent may enumerate/attempt many tools unless the MCP client/agent is constrained.
- ⚠ Apps server requires explicit app selection (--apps); ensure you pass correct app identifiers to avoid missing tools.
- ⚠ The README does not describe MCP-level error codes or retry/backoff guidance; agents may need to rely on MCP client behavior for retries.
Alternatives
Full Evaluation Report
Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for aci-mcp.
AI-powered analysis · PDF + markdown · Delivered within 30 minutes
Package Brief
Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.
Delivered within 10 minutes
Score Monitoring
Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.
Continuous monitoring
Scores are editorial opinions as of 2026-03-30.