adk-deployment-specialist
tessl i github:jeremylongshore/claude-code-plugins-plus-skills --skill adk-deployment-specialistDeploy and orchestrate Vertex AI ADK agents using A2A protocol. Manages AgentCard discovery, task submission, Code Execution Sandbox, and Memory Bank. Use when asked to "deploy ADK agent" or "orchestrate agents". Trigger with phrases like 'deploy', 'infrastructure', or 'CI/CD'.
Review Score
63%
Validation Score
13/16
Implementation Score
42%
Activation Score
77%
Generated
Validation
Total
13/16Score
Passed| Criteria | Score |
|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) |
metadata_version | 'metadata' field is not a dictionary |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata |
Implementation
Suggestions 4
Score
42%Overall Assessment
The skill has good structural organization and appropriate progressive disclosure to external references, but critically lacks actionable, executable content. The instructions read like a high-level checklist rather than concrete guidance Claude can execute. The absence of actual code examples, specific CLI commands, or copy-paste ready configurations significantly limits its utility.
Suggestions
- Add executable code examples for agent definition, including a minimal AgentCard JSON schema and Python agent implementation
- Provide the actual `adk deploy` command with required flags and environment variables, not just a reference to it
- Add explicit validation checkpoints: specific curl commands to test `/.well-known/agent-card` endpoint, expected response schemas, and failure indicators
- Include a concrete example of A2A task submission and status polling with actual request/response payloads
| Dimension | Score | Reasoning |
|---|---|---|
Conciseness | 2/3 | Mostly efficient but includes some unnecessary framing ('Expert in building...') and could be tighter. The overview paragraph explains what ADK is rather than jumping straight to actionable content. |
Actionability | 1/3 | Instructions are abstract and procedural ('Confirm the desired architecture', 'Define the AgentCard') without concrete code, commands, or executable examples. No actual `adk deploy` command syntax or code snippets provided. |
Workflow Clarity | 2/3 | Steps are listed in sequence but lack validation checkpoints. Step 4 mentions 'smoke tests' but doesn't specify what to validate. No explicit feedback loops for deployment failures or rollback procedures. |
Progressive Disclosure | 3/3 | Good structure with clear references to external files for errors and examples. One-level-deep references are well-signaled with explicit paths. Resources section provides relevant external links. |
Activation
Suggestions 2
Score
77%Overall Assessment
The description effectively communicates specific technical capabilities around Vertex AI ADK agents and A2A protocol with good completeness. However, the trigger terms create potential conflict risk by including generic DevOps terminology ('infrastructure', 'CI/CD') that doesn't naturally connect to the specialized agent orchestration focus. The description would benefit from more domain-specific trigger terms.
Suggestions
- Replace generic triggers like 'infrastructure' and 'CI/CD' with more specific terms like 'A2A protocol', 'agent card', 'Vertex AI', or 'ADK' that users would naturally use when needing this specific skill.
- Add file extension or artifact triggers if applicable (e.g., 'agent.yaml', 'AgentCard.json') to improve distinctiveness from general deployment skills.
| Dimension | Score | Reasoning |
|---|---|---|
Specificity | 3/3 | Lists multiple specific concrete actions: 'Deploy and orchestrate Vertex AI ADK agents', 'Manages AgentCard discovery, task submission, Code Execution Sandbox, and Memory Bank'. These are concrete, technical capabilities. |
Completeness | 3/3 | Clearly answers both what ('Deploy and orchestrate Vertex AI ADK agents using A2A protocol. Manages AgentCard discovery, task submission, Code Execution Sandbox, and Memory Bank') and when ('Use when asked to "deploy ADK agent" or "orchestrate agents". Trigger with phrases like...'). |
Trigger Term Quality | 2/3 | Includes some relevant keywords ('deploy', 'infrastructure', 'CI/CD', 'deploy ADK agent', 'orchestrate agents') but the trigger terms feel somewhat disconnected - 'CI/CD' and 'infrastructure' are generic and may not naturally align with the specific ADK/A2A protocol focus. Missing variations like 'A2A', 'agent card', 'vertex'. |
Distinctiveness Conflict Risk | 2/3 | The specific 'Vertex AI ADK' and 'A2A protocol' terminology creates some distinctiveness, but generic triggers like 'deploy', 'infrastructure', and 'CI/CD' could easily conflict with general deployment or DevOps skills. The niche is clear but triggers are too broad. |
- Last updated
- Created
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.