"Run SAM applications, inspect docs/tools, submit tasks through Web
Scanned 9/9/2026
Install to Claude Code
npx -y skills add VectorSpaceLab/AREX-Skill --skill runtime-operations --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Runtime Operations?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/vectorspacelab-runtime-operations)More formats (shields.io, HTML) on the badges page.
---
name: runtime-operations
description: "Run SAM applications, inspect docs/tools, submit tasks through Web
UI or REST gateways, and use the SAM REST client safely."
disable-model-invocation: true
metadata:
disco-role: operating
license: Apache 2.0
---
# runtime-operations
Use this sub-skill when a user needs to operate an existing Solace Agent Mesh (SAM) project: start configured apps, inspect available runtime docs or built-in tools, submit tasks through the Web UI gateway CLI, diagnose gateway reachability, or integrate through the REST gateway and `sam-rest-client`.
## Route first
- Project creation, `sam init`, `sam add`, GUI scaffolding, generated config layout, model-provider setup, artifact/session defaults: route to `../project-bootstrap/SKILL.md`.
- Workflow YAML schema, node wiring, template resolution, DAG validation, branch/map/loop behavior: route to `../workflow-authoring/SKILL.md`.
- `sam eval`, evaluation configs, suites, local/remote evaluation runs: route to `../evaluation/SKILL.md`.
- Plugin creation, install, catalog, and packaging lifecycle: route to `../plugin-lifecycle/SKILL.md`.
## Safe operating stance
Before running anything, decide whether the user asked for a dry inspection or a live runtime action:
- Dry/help/config checks: CLI `--help`, reading generated project files, `sam tools list --json`, and `scripts/check_gateway.py` are safe by default.
- Live execution: `sam run`, `sam task run`, `sam task send`, `sam-rest-cli`, direct REST `POST /api/v2/tasks`, and browser Web UI use may contact brokers, models, gateways, databases, auth services, and user data. Only do these when the user actually asked to run or submit.
- Gateway inspection must not submit real tasks unless the user explicitly asks for a task invocation. The bundled gateway checker only performs GET probes.
## Reference map
- Running projects and Web UI task CLI: `references/running-and-tasks.md`.
- REST gateway client library and CLI: `references/rest-client.md`.
- Built-in docs and tools command family: `references/tool-and-docs-commands.md`.
- Runtime troubleshooting decision tree: `references/troubleshooting.md`.
- Safe gateway checker: `scripts/check_gateway.py`.
## Quick workflows
1. **Start an existing project**: use `references/running-and-tasks.md#run-an-existing-sam-project`; verify config selection and environment loading before live startup.
2. **Send a task to a running Web UI gateway**: use `references/running-and-tasks.md#send-a-task-to-an-already-running-web-ui-gateway`; check agent discovery and session/file options.
3. **One-shot task run**: use `references/running-and-tasks.md#start-sam-send-one-task-and-stop`; distinguish startup timeout from task timeout.
4. **Inspect a gateway without submitting**: run `python scripts/check_gateway.py --url http://localhost:8000 --expect-agent orchestrator` from any directory.
5. **Use REST programmatic access**: use `references/rest-client.md`; install `sam-rest-client` in a separate environment when dependency pins conflict with the main SAM package.
6. **List built-in tools or serve docs**: use `references/tool-and-docs-commands.md`.
## Provenance note
This sub-skill distills behavior from SAM package docs, CLI command source, task/gateway source, REST client source, and unit/integration tests. It is self-contained for runtime use; future agents should not need the original repository checkout to follow these instructions.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!