Terraform Provider Docs

Terraform Provider Documentation Lookup

Why This Skill Exists

Terraform provider APIs have strict argument constraints that are not reliably stored in AI training data. Guessing argument values (port formats, protocol enums, attribute names, etc.) leads to repeated terraform apply failures. Looking up the official docs first saves 5-10 debug cycles.

Real example: alicloud_forward_entry requires port format "1/65535" (slash separator), not "1-65535" (dash). The ip_protocol only accepts any, tcp, udp, forwardgroup — not TCP or UDP. These details are only reliably found in the official docs.

Core Principle

The reason this skill exists is simple: Terraform provider arguments have precise constraints (format strings, enum values, argument interdependencies) that differ between resources even within the same provider. Getting one wrong means a failed terraform apply, reading the error, guessing again, and repeating — often 3-5 times. A single doc lookup before writing the resource eliminates this entire cycle.

Fetch the official documentation page before writing or modifying any Terraform resource or data source. Don't rely on memory or pattern-matching from other resources — each resource has its own unique argument constraints.

How to Look Up Docs

Step 1: Identify the provider and resource type

From the Terraform code, extract:

• Provider: e.g., alicloud, aws, azurerm, google, tencentcloud, volcengine
• Resource/Data source type: e.g., alicloud_forward_entry, aws_instance, google_compute_instance

Step 2: Construct the documentation URL

All Terraform provider docs follow a standard URL pattern on the Terraform Registry:

For resources:

https://registry.terraform.io/providers/{namespace}/{provider}/latest/docs/resources/{resource_name}

For data sources:

https://registry.terraform.io/providers/{namespace}/{provider}/latest/docs/data-sources/{data_source_name}

The {resource_name} is the resource type with the provider prefix stripped. For example, alicloud_forward_entry → forward_entry, aws_instance → instance.

Two quick examples:

• alicloud_nat_gateway → https://registry.terraform.io/providers/aliyun/alicloud/latest/docs/resources/nat_gateway
• aws_security_group → https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/security_group

For the full namespace/provider mapping for all 10 cloud providers in the redc-template project (alicloud, aws, tencent, azure, gcp, volcengine, vultr, huaweicloud, ucloud, ctyun), read references/provider-urls.md.

To generate URLs programmatically, use the bundled helper: python scripts/tf_doc_url.py alicloud_forward_entry

Step 3: Fetch the page

Use web_fetch (or curl) to retrieve the documentation page. Read the Argument Reference section carefully — it lists every argument, its type, whether it's required/optional, and valid values.

What to Look For in the Docs

When you fetch a documentation page, focus on these sections:

1. Argument Reference

• Required vs Optional — which arguments are mandatory
• Type — string, number, list, set, map, block
• Valid values — enums, format patterns, ranges
• Conflicts — arguments that cannot coexist (conflicts_with)
• Force new — arguments that trigger resource recreation

2. Attributes Reference

• What attributes are exported (usable in output blocks or other resources)
• Computed attributes (server-assigned, like id, public_ip)

3. Import

• Whether the resource supports terraform import
• Import ID format

4. Example Usage

• Copy the official example as a starting point
• Pay attention to argument combinations — some are only valid together

When to Look Up Docs

The cost of looking up docs is ~5 seconds. The cost of guessing wrong is a failed terraform apply cycle (30+ seconds per attempt, often 3-5 attempts). The math strongly favors looking things up. Here's when the lookup pays off the most:

1. 🆕 Writing a new resource — even if you've seen similar resources before, argument names and valid values differ
2. 🐛 Debugging a Terraform error — especially InvalidParameter, InvalidValue, Malformed errors that hint at wrong argument format
3. 🔄 Changing arguments — the valid values for one argument may depend on another (e.g., protocol + port constraints)
4. 🤔 Unsure about format — port ranges (1/65535 vs 1-65535), CIDR blocks, protocol names
5. 📊 Using data sources — filter syntax varies wildly between providers
6. 🔗 Referencing attributes — the exported attribute name may differ from what you expect

You can safely skip the lookup when making trivial changes (renaming, changing a string default) or when the resource is already working and you're adjusting a value you're certain about.

Common Pitfalls (Lessons Learned)

These are real issues encountered in the redc-template project. Each was solved instantly by reading the docs:

Workflow Example

Here's the ideal workflow when writing a NAT gateway + DNAT rule for Alibaba Cloud:

1. Need: alicloud_nat_gateway + alicloud_forward_entry

2. Fetch docs:
   → web_fetch("https://registry.terraform.io/providers/aliyun/alicloud/latest/docs/resources/nat_gateway")
   → web_fetch("https://registry.terraform.io/providers/aliyun/alicloud/latest/docs/resources/forward_entry")

3. Read Argument Reference:
   - nat_gateway: nat_type (Enhanced required for EIP), vswitch_id required
   - forward_entry: ip_protocol accepts (any|tcp|udp|forwardgroup)
   - forward_entry: external_port/internal_port format is "port/port" for range
   - forward_entry: port_break required when SNAT shares same EIP

4. Write Terraform code based on docs → works first time ✅

Compare with guessing:

1. Guess port format: "1-65535" → ❌ InvalidExternalPort.Malformed
2. Try "1:65535" → ❌ InvalidExternalPort.Malformed
3. Try port="any" with tcp → ❌ OperationFailed.AnyPortConfig
4. Google the error → find docs → "1/65535" works → ✅ (after 4 attempts)

Bundled Resources

• references/provider-urls.md — Complete URL patterns and examples for all 10 cloud providers. Read this when you need the exact namespace/provider mapping for a provider you haven't used before.
• scripts/tf_doc_url.py — CLI tool to generate doc URLs from resource type names. Run python scripts/tf_doc_url.py <resource_type> to quickly get the URL.