Index of every page on this site: llms.txt
All skills
pulumi avatar

/pulumi-create-agent

@cd1011f official
by pulumipulumi/agent-skills80 stars
8
Run it yourself

Create a reusable Pulumi Agent with saved instructions, a role, and selected integrations. Use when the user asks to create, define, or set up a new Pulumi Agent.

Use this Skill: https://skilld.dev/gh/pulumi/agent-skills/pulumi-create-agent

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencescli-contract.md

≈815 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Agent creation API

Discover operations

Check pulumi version and pulumi api --help. Confirm the backend and account with pulumi whoami -v; if authentication is missing, resolve it before reading organization data.

Use pulumi api list to discover operations and pulumi api describe OPERATION --output=json to inspect their parameters and response fields. Refresh the catalog once with --refresh-spec if an operation is missing.

Operation Purpose
ListAgentDefinitions Find existing names and verify active roster membership
CreateAgentDefinition Create the saved agent
GetAgentDefinition Read the returned ID and verify saved settings
ListRolesByOrgIDAndUXPurpose Read organization roles with uxPurpose=role
GetRole Check the selected role's grants using roleID
ListAgentIntegrations Find connected MCP integrations

Use explicit orgName parameters. Follow pagination where present. Exclude built-in Admin and Billing Manager roles and verify the caller may assign the selected role; holding a role alone does not establish assignment authority.

Select integrations

Read ListAgentIntegrations for the organization. Its integrations array includes unconnected catalog entries too: offer only entries with configured: true.

Map selected integrationId values to enabledIntegrations: [{"id": "SELECTED_ID"}]. Set enabledIntegrations: [] for no MCP integrations. Omitting the field inherits all organization integrations. Credential-field labels and help text describe connection setup; they do not mean an integration is connected.

This workflow selects existing connections. It does not configure new ones.

If integration discovery is unavailable

If discovery, describe, or the integration read fails, stop before creation and save the configuration as a local draft. Preserve the requested selection, identify the failed operation, and explain what is needed to continue. A missing operation, denied access, and a timeout are different failures; none means the organization has no integrations.

Do not substitute internal endpoints, guessed IDs, all integrations, or an empty selection. Resume from verification when the API or access is available.

Build and submit the request

Use the described schema to map the proposal to name, displayName, purpose, customInstructions, roleId, enabledIntegrations, and approvalMode. Check accepted values and limits. Obtain missing team instructions from the user when no supported read exists.

Write the exact displayed values to a JSON file with a JSON serializer, then pass it through --input:

pulumi api CreateAgentDefinition -F orgName=ORG --input agent.json --dry-run --output=json

Replace ORG with the verified organization. Compare the dry-run body with the proposal, then omit --dry-run to create the agent. Dry-run checks request construction, not server authorization.

Read the response using the described response shape, retain the returned agent ID, and verify it with GetAgentDefinition and the active roster. A GET alone may return a deleted definition. Follow the creation skill's recovery guidance if the request times out or saved values differ.

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No alerts3d3 checks · Risk SAFE
  • Gen Agent Trust Hub3d

    This skill is a configuration assistant for creating Pulumi Agents and automation tasks via the Pulumi CLI. It uses standard CLI commands and JSON request files to manage agent definitions and schedules. No security risks were detected.

  • Socket3d

    No alerts

  • Snyk3d

    Risk: LOW · No issues

Signed by skilld at cd1011f. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 4 days ago

README badge

README badge for pulumi/agent-skills/pulumi-create-agent