AGENT SKILL
diagnostics
Run deployment health checks (env, runtime, computation) and return JSON status.
- License 许可
- MIT
- 最近更新
- 2026-11-12
- Stars
- ★ 3
- Forks
- 2
01 前言 Front Matter
- name
- diagnostics
- description
- Run deployment health checks (env, runtime, computation) and return JSON status.
02 资源文件 3
- 脚本__init__.py33 B
- 脚本diagnostics.py3.6 KB
- 参考文档README.md5.8 KB
03 SKILL.md 正文
diagnostics
Verify the agent harness, shell allowlist, and Python runtime are wired correctly.
When to use
- After first deploy or cold start
- When the user asks whether the agent environment is healthy
How to run
Shell commands start in workspace root. Trusted skill files live in the
separate, read-only SKILLS_PATH; use that environment variable when invoking
a bundled script.
list_skillsshould includediagnostics.- Run the script with
run_command:
{"argv": ["bash", "-c", "python3 \"$SKILLS_PATH/diagnostics/diagnostics.py\""]}
- Parse the JSON stdout;
statusshould behealthywhen required env vars are set.
Expected env vars (full check)
AGENT_NAME, VERTEX_AI_PROJECT_ID, SKILLS_DIR — missing vars appear in issues.
04 同仓库更多 Skills
browser
Smoke-test browser skill fixture.
★ 3
plain
暂无简介
★ 3
diagnostics
Inspect a failing run and report the first error.
★ 3
nested
Nested skills must not be discovered.
★ 3
image-generator
Generate images with Gemini image models (Nano Banana Pro) and display them in chat.
★ 3
browser
Control a real browser via CDP for web tasks; check and write site playbooks before improvising.
★ 3