Cloud Infrastructure
aiskillstore/marketplace
Cloud infrastructure design and deployment patterns for AWS, Azure, and GCP.
Designs GCP infrastructure as local Terraform, validates and scans it against best practices, then imports it to Application Design Center for deployment and troubleshooting.
$ npx skills add google/skills --skill application-design-center-design-deploy -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install google/skills application-design-center-design-deploy --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .claude/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .claude/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deployType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add google/skills --skill application-design-center-design-deploy -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install google/skills application-design-center-design-deploy --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .agents/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .agents/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add google/skills --skill application-design-center-design-deploy -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install google/skills application-design-center-design-deploy --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .cursor/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .cursor/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/google/skills.git --path skills/cloud/application-design-center-design-deploy--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add google/skills --skill application-design-center-design-deploy -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install google/skills application-design-center-design-deploy --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .gemini/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .gemini/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install google/skills application-design-center-design-deployInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add google/skills --skill application-design-center-design-deploy -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .github/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .github/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add google/skills --skill application-design-center-design-deploy -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install google/skills application-design-center-design-deploy --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/google/skills.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/cloud/application-design-center-design-deploy .opencode/skills/application-design-center-design-deploy && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "application-design-center-design-deploy" agent skill from https://github.com/google/skills/tree/main/skills/cloud/application-design-center-design-deploy into .opencode/skills/application-design-center-design-deploy/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "application-design-center-design-deploy", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
application-design-center-design-deployDesigns GCP infrastructure as local Terraform, validates and scans it against best practices, then imports it to Application Design Center for deployment and troubleshooting.
The skill lays out a production-minded workflow for the whole infrastructure lifecycle on Google Cloud. It replaces an automated, opaque design tool with a loop the agent controls: modular Terraform written locally, validated with local CLI tools, then given a best-practices plan scan before anything is synchronized with the Application Design Center registry. The local Terraform configuration stays the source of truth, and the agent takes the stance of a principal cloud architect.
Setup comes first: the agent confirms the target project ID and location, defaults to `us-central1` when none is given, and checks that `gcloud` has the right project active. Six phases follow: local design and validation, the plan scan with iterative remediation, import into Application Design Center, deployment and monitoring, troubleshooting failures, and verification with end-to-end tests. The folder has Python scripts to list and fetch Terraform templates, with tests, plus reference guides for planning, generating, validating, remediation and error analysis. It applies only to GCP and to Terraform used within Application Design Center.
6 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 7d97937. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships 4 files in scripts/ (Python), which the agent can run.
Shell commands in SKILL.md call:
terraformgcloudFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use gcloud, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
GCP Application Design Center Deploy loads about 4.4k tokens when it runs, and up to ~18k if it reads all its reference files. Until then it costs about 125 tokens; SKILL.md has 1,815 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
The full file from google/skills at commit 7d97937, republished under its Apache-2.0 licence (© google). 1,815 words, ~4,405 tokens.
.claude/skills/application-design-center-design-deploy/SKILL.md (or your agent's skills folder). This skill also uses 13 other files; get the full folder from GitHub.This skill provides a prescriptive, production-grade workflow for the entire
infrastructure lifecycle on Google Cloud Platform (GCP). It replaces the
automated, opaque-box GAD design_infra tool with an agent-controlled design
and validation loop utilizing modular Terraform and local CLI validation,
followed by a shifted-left best practices plan scan prior to synchronization
with the Application Design Center (ADC) registry for deployment and lifecycle
management.
Always maintain the persona of a Principal Cloud Architect. Keep the local Terraform configuration as the source of truth, and ensure the design is fully compliant with best practices before importing it into the cloud registry.
Before executing Phase 1, you must perform the following setup steps:
Confirm Target Project & Location:
Explicitly ask the user to confirm the target GCP project ID and location (region).
If the user does not specify a location, use us-central1 as the
default.
Verify that your local environment has the active project set:
gcloud config set project <project_id>Goal: Transform user requirements and codebase characteristics into a 100% validated, secure, and compile-ready Terraform configuration locally.
Invoke the design Skill: Call and execute the design skill (defined
in design)
for the user's prompt.
design skill will autonomously perform the Codebase Analysis,
query the catalog registry, planning, HCL generation, and local CLI
validation loop (terraform init, validate, plan) in a dedicated
scratch directory.Locate Validated HCL: Identify the scratch directory where the design
skill saved the validated, compile-ready Terraform files (e.g.,
scratch/tf_validate_<session_id>/).
Verify Handover (MANDATORY): Ensure that the local validation loop in
the design skill completed successfully with a clean plan before
proceeding. Meticulously inspect the HCL to verify:
terraform.tfvars or HCL
resource blocks. All sensitive inputs must be wired through GCP Secret
Manager.backend "gcs" {}) in the HCL files. State must remain
local in the scratch folder during validation, allowing ADC to handle
the remote state registry upon import.Export Terraform Plan to JSON (MANDATORY): In the scratch directory, run the following commands to generate a binary plan and convert it into a clean JSON representation:
terraform plan -out=tfplan && terraform show -json tfplan > tfplan.jsonVerify that the tfplan.json file is successfully written in your scratch
directory.
Goal: Validate the local plan's alignment with security, cost, and reliability benchmarks BEFORE importing it into the cloud registry, using the native ADC plan assessment API.
Discover Space ID (MANDATORY): Before running the assessment or creating templates, you must dynamically discover the active ADC Space ID in your target location:
List Spaces: Run the command:
gcloud design-center spaces list --project=<project_id> --location=<location>Select Space: Parse the output to identify the active space (e.g.,
test-deploy or googlespace). If multiple spaces exist, ask the user
to confirm. If no space exists, ask the user or create one:
gcloud design-center spaces create <space_id> --project=<project_id> --location=<location>Execute Plan Assessment via gcloud: Run the plan-based assessment using
the discovered Space ID and your exported tfplan.json file. Execute the
command directly in your terminal:
gcloud design-center spaces generate-terraform-assessment-report <space_id> \
--location=<location> \
--project=<project_id> \
--terraform-plan="<scratch_directory_path>/tfplan.json" \
--format=jsonAnalyze Findings: Present all findings to the user in a clean tabular format, detailing specific violations, resource scopes, and associated severity levels.
Local Remediation Loop:
Do not attempt to import or commit insecure code.
Edit your local HCL files in the scratch directory to fix the reported violations (e.g., adding encryption keys, enabling OS Login, or restricting IAM scopes).
Re-run Phase 1 local validation and plan export:
terraform validate && terraform plan -out=tfplan && terraform show -json tfplan > tfplan.jsonRe-run the plan assessment command shown in step 2.
Exit Criteria:
Goal: Synchronize the fully validated and best-practice-compliant local HCL configuration with the ADC cloud registry to establish the deployable template resource.
Verify or Create the Application Template (MANDATORY): Before importing the HCL, you must ensure the parent Application Template resource exists in the discovered ADC space.
Check Existence: Run gcloud design-center spaces application-templates describe <template_id> --space=<space_id> --project=<project_id> --location=<location> to check if the template
exists.
Create if Missing: If the describe command returns a NOT_FOUND
error, create the template resource first by running:
gcloud design-center spaces application-templates create <template_id> --space=<space_id> --project=<project_id> --location=<location> --display-name="<Name>" --description="<Description>"Strict HCL Parser Constraints (CRITICAL): Before calling the import operation, ensure your local HCL complies with the ADC registry's strict ingestion rules:
resource blocks inside the imported HCL. Only
module, variable, output, and provider blocks are allowed. If a
resource is required (e.g. Private Service Access peering) but no
standalone module is registered for it in the catalog, you MUST check if
it is supported as a built-in configuration option inside an existing
registered module (e.g. setting private_service_access_config inside
module "vpc").subnet_private_access = "true", NOT
as a boolean true.terraform {}
version constraint block. Omit it entirely from providers.tf or
main.tf.Import to ADC Template: Once the template resource is confirmed to exist
and the HCL is validated against the above constraints, invoke the hosted
application_design_center:manage_application_template MCP tool with the
APPLICATION_TEMPLATE_OPERATION_IMPORT_IAC operation:
Arguments:
project: The target project ID.
location: The GCP deployment region (e.g., us-central1).
spaceId: The discovered ADC space ID.
applicationTemplateId: A unique name for your application
template.
operation: APPLICATION_TEMPLATE_OPERATION_IMPORT_IAC
iacModule: A structured object containing the files list:
{
"files": [
{ "name": "main.tf", "content": "<content of main.tf>" },
{ "name": "variables.tf", "content": "<content of variables.tf>" },
{ "name": "terraform.tfvars", "content": "<content of terraform.tfvars>" }
]
}Resilience & Retries (MANDATORY):
IMPORT_IAC call fails due to a transient error (e.g., 502 Bad Gateway, 504 Gateway Timeout, or 429 Rate Limit), do not
immediately retry.gcloud alpha design-center spaces application-templates describe
to check if the import actually succeeded in the background. Only
retry if the template was not updated.Capture Template URI: Upon success, this establishes the template
resource in your space. Construct the applicationTemplateUri using the
pattern:
projects/{project}/locations/{location}/spaces/{spaceId}/applicationTemplates/{applicationTemplateId}
Goal: Deploy the validated, best-practice-compliant application template to the GCP environment.
application_design_center:manage_application MCP tool with the
APPLICATION_OPERATION_DEPLOY operation:project: Target project ID.location: Target deployment location.spaceId: Target space ID.applicationId: A unique ID for the deployed application instance.applicationTemplateUri: The URI established in Phase 3.serviceAccount: The deployment service account.DEPLOY operation fails with transient network or gateway
errors (e.g., 502, 504), apply exponential backoff with
jitter before retrying.gcloud design-center spaces applications describe to confirm its status before retrying the
deploy call, avoiding concurrent conflicting deployments.done: true using the command gcloud design-center operations describe <operation_name>.done is true and there is no error field, proceed
to Phase 6.error field is present, analyze the error type and
proceed to Phase 5.Goal: Diagnose and remediate deployment failures iteratively using the specialized troubleshooting skill and established cloud resolution patterns.
Iterative Cloud Resolution Patterns (CRITICAL): If the deployment fails
with a REVISION_FAILED or TERRAFORM error, check for these common
resource conflicts:
Service Account 409 Conflict (alreadyExists): If the deployment
fails because a service account generated by the module (e.g.
frontend-service-us-central-sa) already exists in the project,
remediate the local HCL by disabling service account creation and
referencing the existing one:
create_service_account = false
service_account = "<existing_service_account_email>"Container Image 404 NotFound: If the deployment fails because a
container image is not found, confirm that the image exists in your
registry. For testing or hello-world deployments, leverage the official
public Google hello-world image:
us-docker.pkg.dev/cloudrun/container/hello
Delegate to the Troubleshooting Skill: If a deployment failure occurs
and does not match the above patterns, invoke and execute the specialized
infra-deployment-debugging guide (located in
infra-deployment-debugging).
Select the Troubleshooting Context:
Apply Local-First Remediation:
Follow the troubleshooting skill's remediation guides to formulate a fix.
MANDATORY: Apply the fix directly to your local HCL files in the scratch directory, re-run local validation, re-import the HCL, and trigger a new deployment.
Re-run Phase 1 local validation and plan export:
terraform validate && terraform plan -out=tfplan && terraform show -json tfplan > tfplan.jsonRe-run the plan assessment (Phase 2) to ensure no new violations are introduced.
Re-import the corrected HCL to ADC using
APPLICATION_TEMPLATE_OPERATION_IMPORT_IAC.
Trigger a new deployment using APPLICATION_OPERATION_DEPLOY.
Iteration Threshold: Repeat the troubleshooting, validation, import, and redeployment cycle up to five (5) times. If it still fails, report the full history and diagnostics to the user.
Goal: Confirm that the deployed services are healthy and fully functional.
application_design_center:manage_application MCP tool with the
APPLICATION_OPERATION_GET operation to retrieve the resource details,
public endpoints, and output parameters.Report bugs or improvements for this skill at Google Skills Issues.
© google, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 13 other files (scripts, references) in skills/cloud/application-design-center-design-deploy of google/skills.
Open the folder on GitHubat commit 7d97937
GCP Application Design Center Deploy next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| GCP Application Design Center Deploy this skillgoogle/skills | 21k | — | ~4.4k | Automated safety check: Pass | Apache-2.0 | |
| Cloud Infrastructureaiskillstore/marketplace | 430 | 1 repos | ~1.3k | Automated safety check: Pass | None | |
| Senior DevOps Toolkitmaslennikov-ig/claude-code-orchestrator-kit | 260 | 6 repos | ~1.1k | Automated safety check: Notes | Custom licence | |
| Terraform Module Librarywshobson/agents | 40k | 11 repos | ~1.3k | Automated safety check: Pass | MIT | |
| DeployingGoogleCloudPlatform/race-condition | 234 | — | ~3k | Automated safety check: Pass | Custom licence | |
| Cloud Architectdavila7/claude-code-templates | 32k | 8 repos | ~1.9k | Automated safety check: Pass | MIT |
aiskillstore/marketplace
Cloud infrastructure design and deployment patterns for AWS, Azure, and GCP.
maslennikov-ig/claude-code-orchestrator-kit
Comprehensive DevOps skill for CI/CD, infrastructure automation, containerization, and cloud platforms (AWS, GCP, Azure). Includes pipeline setup…
wshobson/agents
Build reusable, tested Terraform modules for AWS, Azure, GCP and OCI, with a standard file layout, an AWS VPC example, versioning rules and Terratest checks.
GoogleCloudPlatform/race-condition
Guides deployment of Race Condition to a GCP project. An agent skill from GoogleCloudPlatform/race-condition.
davila7/claude-code-templates
Expert cloud architect specializing in AWS/Azure/GCP multi-cloud infrastructure design, advanced IaC (Terraform/OpenTofu/CDK), FinOps cost optimization, and modern architectural patterns.
Jeffallan/claude-skills
Writes reusable Terraform modules and manages state, providers and environments across AWS, Azure and GCP, with validation, plan review and explicit apply approval.
google/skills
Query Cloud Trace spans, filter by latency thresholds or error status, correlate distributed traces with Cloud Logging, and diagnose latency bottlenecks across Google Cloud services.
google/skills
Manages Google Cloud Privileged Access Manager entitlements and grants: create and edit entitlements, request temporary access, and approve or deny pending grants.
google/skills
Writes Terraform alerting policies for AI agents that emit OpenTelemetry metrics, covering reliability, cost, safety, security and quality signals on Google Cloud.
google/skills
Deploys open models or custom weights from Model Garden to Agent Platform endpoints, checks deployment status and cleans up endpoints, confirming before any change.
google/skills
Searches, manages and scaffolds skills in the Gemini Enterprise Agent Platform Skill Registry using bundled Python scripts and Google Cloud credentials.
google/skills
Analyzes BigQuery slot use, query costs and execution bottlenecks from INFORMATION_SCHEMA to diagnose slow queries, slot contention and unpartitioned scans.
Works with
Categories
Designs GCP infrastructure as local Terraform, validates and scans it against best practices, then imports it to Application Design Center for deployment and troubleshooting. The skill lays out a production-minded workflow for the whole infrastructure lifecycle on Google Cloud. It replaces an automated, opaque design tool with a loop the agent controls: modular Terraform written locally, validated with local CLI tools, then given a best-practices plan scan before anything is synchronized with the Application Design Center registry.
GCP Application Design Center Deploy fits situations like: designing GCP infrastructure with Terraform from a set of requirements; validating local HCL and scanning the plan before deploying; importing a Terraform template into Application Design Center and deploying it; troubleshooting a failed Application Design Center deployment.
Run `npx skills add google/skills --skill application-design-center-design-deploy -a claude-code`. Or copy the skill folder (skills/cloud/application-design-center-design-deploy in google/skills) into .claude/skills/application-design-center-design-deploy in your project. Claude Code loads it when a task matches its description.
Run `npx skills add google/skills --skill application-design-center-design-deploy -a codex`. Or copy the skill folder (skills/cloud/application-design-center-design-deploy in google/skills) into .agents/skills/application-design-center-design-deploy in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add google/skills --skill application-design-center-design-deploy -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/application-design-center-design-deploy, .gemini/skills/application-design-center-design-deploy, .github/skills/application-design-center-design-deploy and .opencode/skills/application-design-center-design-deploy in your project.
Going by SKILL.md and its folder, GCP Application Design Center Deploy needs Python for the scripts in its folder and the command-line tools its instructions call (terraform and gcloud). Our summary lists: A GCP project ID and location; gcloud CLI with an active project; Terraform for local validation.
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
GCP Application Design Center Deploy is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.4k tokens (SKILL.md is roughly 18k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 13k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with GCP Application Design Center Deploy: Cloud Infrastructure (aiskillstore/marketplace, 430 stars), Senior DevOps Toolkit (maslennikov-ig/claude-code-orchestrator-kit, 260 stars), Terraform Module Library (wshobson/agents, 40k stars) and Deploying (GoogleCloudPlatform/race-condition, 234 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
google (a GitHub organization, an official publisher) maintains it in google/skills, which has 21,032 GitHub stars. The repository holds 147 skills in this directory. The repository was last updated on October 8, 2026.
Source: google/skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.