菜单
Manage the hal-health sidecar container that serves live ecosystem state snapshots to HAL Plus and other consumers. Use when the user asks to refresh the status snapshot, recreate the health container, or troubleshoot HAL Plus product state.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Deploy the local Terraform demo environment in hal. Use to provision and initialize Terraform lab resources.
Route Terraform lab requests to deploy, destroy, inspect status, and manage workspace automation in hal.
Deploy local Vault in hal. Use when the user asks to start Vault, initialize the local control plane, or recover from Vault being offline.
Route user intent to the right Vault lab skill in hal, including CE versus Enterprise constraints, observability wiring, and full-stack scenario guidance.
Deploy, verify, and troubleshoot the Vault JWT auth lab in hal. Use this skill when the user asks to enable JWT auth, configure CI/CD authentication with Vault, debug bound claims, test GitLab or pipeline logins, or reset the JWT demo. Triggers include "enable jwt", "configure jwt", "gitlab vault auth", "pipeline auth", "bound claims", and "hal vault jwt".
Deploy, verify, and troubleshoot the Vault OIDC lab in hal. Use this skill when the user asks to enable OIDC, set up human SSO, debug browser login redirects, inspect Authentik-backed roles, reset the OIDC demo, or configure SCIM provisioning. Triggers include "enable oidc", "Vault SSO", "Authentik", "OIDC callback", "configure oidc", "hal vault oidc", "enable scim", and "vault scim".
| name | health |
| description | Manage the hal-health sidecar container that serves live ecosystem state snapshots to HAL Plus and other consumers. Use when the user asks to refresh the status snapshot, recreate the health container, or troubleshoot HAL Plus product state. |
The hal health namespace manages the hal-health sidecar container. It builds a frozen snapshot of the running HAL ecosystem and serves it at http://hal-health:9001/api/status on hal-net so HAL Plus and AI clients can query product state without touching the container engine directly.
hal health create # create (or recreate) the hal-health container with a fresh snapshot
hal health update # refresh the snapshot — use after ecosystem changes outside normal lifecycle
hal health delete # remove the hal-health container
hal health create: first-time setup or after hal health delete.hal health update: manual escape hatch after a product was added/removed outside the normal hal <product> create/delete flow (e.g. manual container changes).hal health delete: clean up the sidecar without affecting any product containers.hal-health container reuses ghcr.io/hashimiche/hal-mcp:latest with --entrypoint /usr/local/bin/hal and health _serve as args.{ timestamp, engine, products: [{ product, state, health, reason, endpoint, containers, features: [...] }] }.HAL_HEALTH_DATA env var injected at start time and never touches the container engine socket.global.RefreshHalHealth is called automatically after all product create/delete commands — manual hal health update is only needed for out-of-band changes.hal health is a no-op if hal-net does not exist or ghcr.io/hashimiche/hal-mcp:latest is not present locally.hal-net for hal-health to be reachable by HAL Plus.hal-health is unavailable.