Gemini Enterprise Skill Registry
Search, manage, monitor, and scaffold skills for the Gemini Enterprise Agent Platform.
The evidence shows use of Google Cloud default credentials and Bearer-token API calls, with read-only discovery distinguished from lifecycle changes. However, upload, update, and permanent deletion lack user confirmation, least-privilege guidance, sensitive-data-flow disclosure, and rollback procedures, so points were deducted.
The script implements the main documented commands and checks project and location environment variables. However, it lacks timeouts, retries, response-schema validation, partial-failure recovery, and tests; the mandatory validation step is documented but not enforced by the main script, so points were deducted.
The audience and scenarios are reasonably clear, covering search, retrieval, revision inspection, lifecycle management, and skill scaffolding. However, the main skill lacks complete use/non-use boundaries, input/output contracts, and trigger conditions; Chinese support and reachability of Google Cloud services from mainland China are not addressed, so points were deducted.
Documentation is layered into capabilities, prerequisites, quickstart, and references, with useful command examples; the repository also provides Apache-2.0 licensing, installation guidance, and an issue path. However, there is no skill version, changelog, explicit maintenance owner, or dependency pinning, so points were deducted.
The skill provides directly reusable commands for discovery, retrieval, upload, update, deletion, revision inspection, and operation monitoring, covering the core registry workflow. However, results still require user interpretation, generated skills are explicitly only drafts, there is no representative execution evidence, and abnormal-case handling is limited, so points were deducted.
The repository includes a source implementation corresponding to the documented commands and an auditable environment-validation script. However, there is no committed test suite, CI coverage, third-party execution evidence, or corroborating source material; static review cannot verify API behavior, so points were deducted.
- Deletion is documented as permanent but has no confirmation, backup, or recovery workflow; manually verify the project, location, and skill ID before execution.
- The script reads local files, packages directories, and uploads them to Google Cloud; inspect archive contents to avoid sending credentials or sensitive data.
- Dependencies and Google Cloud API behavior are not pinned or statically verified; begin with read-only operations in an isolated environment and validate against current API documentation.
- Reachability from mainland-China networks is not documented, so additional network configuration may be required.
What does this skill do, and when should you use it?
This skill provides instructions for interacting with the Gemini Enterprise Agent Platform Skill Registry. It supports skill discovery, including searching, listing, retrieving specific skills, and inspecting revision histories. It also covers uploading, updating, permanently deleting skills, monitoring long-running operations, and generating an initial local skill scaffold. Use requires environment validation, Google Cloud authentication, and project and location configuration.
Runs python3 scripts/validate_env.py before registry operations; uses scripts/skill_registry_ops.py to search for skills; provides procedures for querying, listing, retrieving skills, and inspecting revision histories; uploads, updates, or permanently deletes skills; checks completion status for long-running operations; and generates an initial skill scaffold locally.
- An agent developer looking for existing capabilities can search and inspect skills in the Gemini Enterprise registry.
- A team publishing a new skill can upload it and monitor the associated long-running operation.
- An administrator maintaining registered skills can update or permanently delete them.
- An engineer starting a new agent skill can generate an initial local scaffold.
What are this skill's strengths and limitations?
- Covers discovery, lifecycle management, long-running operation monitoring, and skill scaffolding.
- States clear validation and Google Cloud authentication prerequisites.
- Can be selected independently through the repository installation flow.
- Requires Google Cloud credentials, a project, and a location.
- The source does not provide complete commands for upload, update, deletion, or scaffolding.
- The source provides no test-suite evidence, platform validation results, or pricing information.
- Permanent deletion is an irreversible lifecycle operation and requires care.
How do you install this skill?
Run npx skills add google/skills and select agent-platform-skill-registry during installation. The source does not specify the exact destination folder after selection. Before use, install dependencies with pip install google-auth requests, then authenticate with gcloud auth application-default login.
How do you use this skill?
First run python3 scripts/validate_env.py. Set GCP_PROJECT_ID and GCP_LOCATION, and make sure Google Cloud credentials are available. Search example: python3 scripts/skill_registry_ops.py search --query "test skill" --top-k 5. For the remaining discovery, lifecycle, monitoring, and scaffolding commands, consult the corresponding documents in the skill directory; their complete syntax is not included in the source material.