Cor-IncorporatedCor-Incorporated/claude-code-skills/skills/adk-engineer/SKILL.md

adk-engineer

Skill5 starsChanged 2 months agoMITView on GitHub

Implement production-ready agents using Google Agent Development Kit (ADK). Use when creating ADK agents, building multi-agent systems, implementing workflow agents, reviewing ADK code, or deploying to Vertex AI / Cloud Run. Triggers: 'ADK agent', 'Google ADK', 'multi-agent system', 'Gemini agent', 'agent workflow', 'adk deploy'. Do NOT use for non-ADK agent frameworks (LangChain, CrewAI, AutoGen) or general Gemini API calls without ADK.

1 other project uses a version of this.

Adopted by

Most recently taken up by

RuslanKhis/agentic-engineering-skillstheir own version, first seen Oct 27, 2026

Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
RuslanKhis/agentic-engineering-skills★ 9same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository25 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-orchestratorSkill
- brainstormingSkill
- bugfixSkill
- classify-reviewSkill
- code-reviewerSkill
- codex-reviewSkill
- context7-skillsSkill
- developer-growth-analysisSkill
- file-organizerSkill
- gcp-deploy-guardianSkill
- gws-workspaceSkill
- handoverSkill
- modern-architectureSkill
- planSkill
- repo-cleanSkill
- review-loopSkill
- security-reviewSkill
- skill-creatorSkill
- supabase-nextjs-debuggerSkill
- tdd-workflowSkill
- test-falsifySkill
- ui-design-systemSkill
- ui-skillsSkill
- ux-researcher-designerSkill

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Single agent: Read references/adk-patterns.md -> Use assets/templates/single-agent/ -> Validate with references/adk-best-practices.md.
line 13
Very specific:
Test locally: Run python agent.py or adk ui.
line 46
Distinctive:
Read agent code and identify patterns used.
line 59
Very specific:
Check implementation against references/adk-patterns.md.
line 60
Very specific:
If validate_agent.py reports issues, fix all CRITICAL/HIGH items before deployment.
line 103
Distinctive:
If pattern selection is ambiguous, default to Coordinator for routing tasks, Sequential for pipelines.
line 104
Distinctive:
If deployment fails, check project permissions and staging bucket access first.
line 105

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

ADK Engineer

Implement, review, and deploy Google ADK agents following proven patterns and best practices.

Quick Start

Single agent: Read references/adk-patterns.md -> Use assets/templates/single-agent/ -> Validate with references/adk-best-practices.md.

Multi-agent system: Choose pattern (Coordinator, Sequential, Parallel, Loop) -> Use matching template from assets/templates/ -> Consult references/adk-patterns.md.

<important if="choosing an ADK agent pattern or designing a multi-agent system architecture">

Pattern Selection

Task Type	Pattern	Template
Linear data processing	Sequential Pipeline	`assets/templates/workflow/`
Intent-based routing	Coordinator/Dispatcher	`assets/templates/multi-agent/`
Independent concurrent tasks	Parallel Fan-Out/Gather	`assets/templates/workflow/`
Nested agent structures	Hierarchical Decomposition	`assets/templates/multi-agent/`
Validation loops	Generator-Critic	`assets/templates/workflow/`
Quality improvement	Iterative Refinement	`assets/templates/workflow/`
Critical decisions	Human-in-the-Loop	`assets/templates/single-agent/`

</important>

Implementation Workflow

<important if="creating a new ADK agent from a template">

New Agent

Choose pattern from table above based on task analysis.

Copy template:

cp -r assets/templates/[single-agent|multi-agent|workflow]/ ./my_agent/
# Or: bash scripts/init_adk_project.sh my_agent [template_type]

Implement domain logic: Define custom tools, update instructions, configure output_key values.
Test locally: Run python agent.py or adk ui.

Validate:

python scripts/validate_agent.py agent.py

Deploy per references/adk-deployment.md.

</important><important if="reviewing existing ADK agent code for correctness and best practices">

Code Review

Read agent code and identify patterns used.
Check implementation against references/adk-patterns.md.
Review against references/adk-best-practices.md.
Run python scripts/validate_agent.py agent.py.
Provide specific, actionable recommendations with references.

</important><important if="validating an ADK agent before deployment or during code review">

Validation Checklist

root_agent properly defined
Agent descriptions are clear and specific
Descriptive output_key values for state management
Loop agents have max_iterations
Parallel agents use unique output_key values
Tools follow verb-noun naming pattern
Error handling with graceful degradation
Appropriate model selection per agent complexity
REQUIRE_CONFIRMATION for high-stakes operations

</important><important if="deploying an ADK agent to Vertex AI Agent Engine or Cloud Run">

Deployment

# Vertex AI Agent Engine (recommended)
adk deploy agent_engine --project=PROJECT_ID --region=us-central1 \
  --staging_bucket=gs://BUCKET/staging --agent-file=agent.py

# Cloud Run
adk deploy cloud_run --project=PROJECT_ID --region=us-central1 \
  --agent-file=agent.py

See references/adk-deployment.md for Docker, CI/CD, monitoring, and security details.

</important>

Error Handling

If validate_agent.py reports issues, fix all CRITICAL/HIGH items before deployment.
If pattern selection is ambiguous, default to Coordinator for routing tasks, Sequential for pipelines.
If deployment fails, check project permissions and staging bucket access first.

References

references/adk-patterns.md -- 7 design patterns with code examples
references/adk-best-practices.md -- Agent design, state management, tools, security
references/adk-deployment.md -- Deployment workflows, CI/CD, monitoring, cost optimization
assets/templates/ -- Single-agent, multi-agent, workflow starter templates
Google ADK Docs

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.

Check your setup, free