agentleFS
pr-pmpr-pm/prpm/.claude/skills/pulumi-troubleshooting/SKILL.md

pulumi-troubleshooting

Skill124 starsChanged 8 months agoMITView on GitHub

Comprehensive guide to troubleshooting Pulumi TypeScript errors, infrastructure issues, and best practices - covers common errors, Outputs handling, AWS Beanstalk deployment, and cost optimization

  • Reads credentials

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 8 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Solution: Use optional chaining for conditional exports:

    line 29
  2. Distinctive:

    Solution: Remove the config section or don't set defaults for provider configs:

    line 44
  3. Very specific:

    Solution: Use || operator to create if not exists:

    line 62
  4. Very specific:

    Use .apply() to transform Output values

    line 71
  5. Very specific:

    Use pulumi.all([...]) to combine multiple Outputs

    line 72
  6. Very specific:

    Use pulumi.output(value) to wrap plain values as Outputs

    line 73
  7. Distinctive:

    Solution: Use JSON.stringify for complex values:

    line 109
  8. Distinctive:

    Solution: Ensure DNS records are created and wait for validation:

    line 122
  9. Very specific:

    Run npm run build in the infra package locally

    line 165
  10. Distinctive:

    Check for conditional exports inside blocks

    line 166
  11. Distinctive:

    Verify all Pulumi Outputs are properly typed

    line 167
  12. Very specific:

    Look for .map() calls on potentially undefined arrays

    line 168
  13. Distinctive:

    Ensure all imports are correct

    line 169
  14. Distinctive:

    Use public subnets to avoid NAT Gateway costs ($32/month)

    line 177

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Pulumi Infrastructure Troubleshooting Skill

Common Pulumi TypeScript Errors and Solutions

1. "This expression is not callable. Type 'never' has no call signatures"

Cause: TypeScript infers a type as never when working with Pulumi Outputs, especially with arrays.

Solution: Wrap the value in pulumi.output() and properly type the callback:

// ❌ Bad - TypeScript can't infer the type
value: pulumi.all(config.vpc.publicSubnets.map((s: any) => s.id))

// ✅ Good - Explicitly wrap and type
value: pulumi.output(config.vpc.publicSubnets).apply((subnets: any[]) =>
  pulumi.all(subnets.map((s: any) => s.id)).apply(ids => ids.join(","))
)

2. "Modifiers cannot appear here" (export in conditional blocks)

Cause: TypeScript doesn't allow export statements inside if blocks.

Solution: Use optional chaining for conditional exports:

// ❌ Bad
if (opensearch) {
  export const opensearchEndpoint = opensearch.endpoint;
}

// ✅ Good
export const opensearchEndpoint = opensearch?.endpoint;

3. "Configuration key 'aws:region' is not namespaced by the project"

Cause: Pulumi.yaml config with namespaced keys (e.g., aws:region) cannot use default attribute.

Solution: Remove the config section or don't set defaults for provider configs:

# ❌ Bad
config:
  aws:region:
    description: AWS region
    default: us-east-1

# ✅ Good - set via workflow/CLI instead
config:
  app:domainName:
    description: Domain name

4. Stack Not Found Errors

Cause: Pulumi stack doesn't exist yet in new environments.

Solution: Use || operator to create if not exists:

pulumi stack select $STACK || pulumi stack init $STACK

5. Working with Pulumi Outputs

Key Concepts:

  • pulumi.Output<T> is a promise-like wrapper for async values
  • Use .apply() to transform Output values
  • Use pulumi.all([...]) to combine multiple Outputs
  • Use pulumi.output(value) to wrap plain values as Outputs

Common Patterns:

// Transforming a single Output
const url = endpoint.apply(e => `https://${e}`);

// Combining multiple Outputs
const connectionString = pulumi.all([host, port, db]).apply(
  ([h, p, d]) => `postgres://${h}:${p}/${d}`
);

// Interpolating Outputs
const message = pulumi.interpolate`Server at ${endpoint}:${port}`;

Nested Outputs (Properties that are themselves Outputs):

// ❌ Bad - resource.property might be an Output<string>
const endpoint = instance.apply(i => i.endpoint.split(":")[0]); // ERROR: Property 'split' does not exist

// ✅ Good - unwrap nested Output with pulumi.output()
const endpoint = instance.apply(i =>
  pulumi.output(i.endpoint).apply(e => e.split(":")[0])
);

// ✅ Alternative - use pulumi.all to flatten
const endpoint = pulumi.all([instance]).apply(([inst]) =>
  pulumi.output(inst.endpoint).apply(e => e.split(":")[0])
);

6. Beanstalk Environment Variables

Issue: Complex objects or arrays need to be serialized.

Solution: Use JSON.stringify for complex values:

{
  namespace: "aws:elasticbeanstalk:application:environment",
  name: "ALLOWED_ORIGINS",
  value: allowedOrigins.apply(origins => JSON.stringify(origins)),
}

7. ACM Certificate Validation

Issue: Certificate validation hangs or times out.

Solution: Ensure DNS records are created and wait for validation:

// 1. Create certificate
const cert = new aws.acm.Certificate(...);

// 2. Create DNS validation record
const validationRecord = new aws.route53.Record(..., {
  name: cert.domainValidationOptions[0].resourceRecordName,
  type: cert.domainValidationOptions[0].resourceRecordType,
  records: [cert.domainValidationOptions[0].resourceRecordValue],
});

// 3. Wait for validation to complete
const validation = new aws.acm.CertificateValidation(..., {
  certificateArn: cert.arn,
  validationRecordFqdns: [validationRecord.fqdn],
});

8. GitHub Actions Pulumi Setup

Best Practices:

- name: Setup Pulumi
  uses: pulumi/actions@v5

- name: Configure Stack
  run: |
    STACK="${{ inputs.stack || 'prod' }}"
    pulumi stack select $STACK || pulumi stack init $STACK
    pulumi config set aws:region ${{ env.AWS_REGION }}
    # Set other non-secret configs here

- name: Pulumi Up
  run: pulumi up --yes --non-interactive
  env:
    PULUMI_ACCESS_TOKEN: ${{ secrets.PULUMI_ACCESS_TOKEN }}
    PULUMI_CONFIG_PASSPHRASE: ${{ secrets.PULUMI_CONFIG_PASSPHRASE }}

9. Debugging TypeScript Compilation

Quick checks:

  1. Run npm run build in the infra package locally
  2. Check for conditional exports inside blocks
  3. Verify all Pulumi Outputs are properly typed
  4. Look for .map() calls on potentially undefined arrays
  5. Ensure all imports are correct

10. Cost Optimization Tips

Beanstalk vs ECS Fargate:

  • Beanstalk with t3.micro: ~$32/month
  • ECS Fargate: ~$126/month
  • Key difference: Beanstalk runs on EC2 instances you control
  • Use public subnets to avoid NAT Gateway costs ($32/month)

Checklist Before Deploying

  • Run npm run build locally to catch TypeScript errors
  • Test with pulumi preview before pulumi up
  • Verify all secrets are in GitHub Secrets (not hardcoded)
  • Check stack name matches environment (dev/staging/prod)
  • Ensure domain/DNS is configured if using custom domains
  • Verify VPC/subnets exist if using existing infrastructure
  • Check that all required extensions/providers are installed

Common Environment Variables to Set

// Database
DATABASE_URL: pulumi.interpolate`postgres://${user}:${pass}@${host}:5432/${db}`

// Redis
REDIS_URL: redisEndpoint.apply(e => `redis://${e}:6379`)

// S3
S3_BUCKET: bucketName
S3_REGION: region

// Auth
GITHUB_CLIENT_ID: clientId
GITHUB_CLIENT_SECRET: clientSecret

// App Config
NODE_ENV: "production"
PORT: "8080"
LOG_LEVEL: "info"

Resources

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.