astro-airflow-mcp

astro-airflow-mcp is a Model Context Protocol (MCP) server (FastMCP-based) that exposes Apache Airflow functionality (via Airflow’s REST API) to AI agents. It can run standalone over MCP (stdio/HTTP) or as an Airflow 3.x plugin, and supports multiple Airflow auth styles (bearer token, basic auth, and username/password with OAuth2 token exchange for Airflow 3.x).

Evaluated Mar 30, 2026 (21d ago)
Homepage ↗ Repo ↗ Automation mcp fastmcp python airflow agents workflow-orchestration
⚙ Agent Friendliness
62
/ 100
Can an agent use this?
🔒 Security
58
/ 100
Is it safe for agents?
⚡ Reliability
31
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
80
Documentation
70
Error Messages
0
Auth Simplicity
65
Rate Limits
20

🔒 Security

TLS Enforcement
70
Auth Strength
70
Scope Granularity
30
Dep. Hygiene
65
Secret Handling
55

The tool explicitly mentions excluding connection credentials from list_connections (positive). However, documentation primarily describes how to supply Airflow credentials via environment variables and does not describe fine-grained authorization/scope enforcement, audit logging, CSRF considerations, or explicit transport security requirements for the MCP server itself.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
40
Error Recovery
40
AF Security Reliability

Best When

You want an agent-friendly, typed MCP interface to Airflow metadata and common operational workflows, especially when you already run Apache Airflow and can provide credentials for its REST API.

Avoid When

You need strict rate-limit/latency guarantees or you cannot provide secure secret handling (e.g., safe credential storage) for agent-originated requests.

Use Cases

  • Let AI agents inspect Airflow DAGs/tasks/runs and retrieve source/logs for troubleshooting
  • Agent-driven health checks (imports/warnings/errors, DAG statistics)
  • Agent-initiated operational actions like triggering or pausing/unpausing DAGs
  • Centralized agent access to Airflow metadata (providers/plugins/config/assets/variables/pools)

Not For

  • Highly regulated environments that require strict auditing/least-privilege controls over operational changes
  • Use cases needing guaranteed idempotency for orchestration actions without additional safeguards
  • Environments where outbound network access to the Airflow webserver is not allowed

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: Bearer token (AIRFLOW_AUTH_TOKEN) Username/password (AIRFLOW_USERNAME/AIRFLOW_PASSWORD) with OAuth2 token exchange for Airflow 3.x Basic auth (Airflow 2.x)
OAuth: Yes Scopes: No

Auth is delegated to Airflow’s REST API auth mechanisms; the MCP server relays credentials/tokens. README indicates Airflow 3.x uses OAuth2 token exchange for username/password.

Pricing

Free tier: No
Requires CC: No

Open-source project (Apache-2.0). Costs depend on your Airflow infrastructure and any MCP client usage.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Mutating tools (e.g., trigger_dag, pause_dag, unpause_dag) can cause real side effects; agents should use explicit confirmation/guardrails.
  • Large log/task outputs may be truncated/limited by the underlying Airflow REST API responses (not documented here).
  • Server defaults to HTTP mode but MCP clients may require stdio transport; misconfiguration can prevent connection.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for astro-airflow-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered