---

name: solace-agent-mesh description: "Use Solace Agent Mesh to scaffold projects, configure agents and gateways, run tasks, author workflows, manage plugins, and evaluate agent systems." disable-model-invocation: true metadata: disco-role: operating license: Apache 2.0

Solace Agent Mesh

Use this repo skill when a task mentions Solace Agent Mesh, SAM, the sam CLI, event-driven agent meshes, SAM projects, agents, gateways, proxies, workflows, plugins, task submission, Web UI/REST gateways, built-in tools, or sam eval.

Before doing live work, decide whether the user wants a dry planning/validation task or a live runtime action. Live actions can start brokers/gateways, call LLM providers, open browsers, write project files, install plugins, submit tasks, or run evaluations.

First checks

• Package runtime: Python >=3.10.16,<3.14; public distribution solace-agent-mesh; CLI entry points sam and solace-agent-mesh.

• Install the main package in the active environment with python -m pip install solace-agent-mesh or, for a source checkout, build/install a normal local wheel instead of relying on editable CLI behavior.

• Optional REST client package: install sam-rest-client separately when you need sam-rest-cli or SAMRestClient. Treat it as a separate install surface when dependency pins conflict with the main package.

• For a safe import/CLI smoke check, run the bundled helper from this skill tree:

  python scripts/check_install.py
  python scripts/check_install.py --include-rest-client      # when both packages are intentionally installed together
  python scripts/check_install.py --rest-client-only        # when sam-rest-client is isolated
  

• Read references/repo-provenance.md before deciding whether this skill is current for a checkout.

• Use references/package-overview.md, references/cli-reference.md, and references/configuration-concepts.md for shared context.

• Use references/troubleshooting.md for cross-cutting install/import/build/service issues.

Route map

User intent	Read this
Initialize a project, plan sam init, add agents/gateways/proxies, inspect generated layouts, or configure shared services	sub-skills/project-bootstrap/SKILL.md
Create, install, catalog, add, inspect, or build SAM plugin packages	sub-skills/plugin-lifecycle/SKILL.md
Start an existing SAM app, inspect built-in tools/docs, submit tasks, diagnose gateways, or use sam-rest-client	sub-skills/runtime-operations/SKILL.md
Author or validate workflow YAML/DAGs, node dependencies, map/switch/loop behavior, or template expressions	sub-skills/workflow-authoring/SKILL.md
Configure sam eval, draft evaluation suites/test cases, validate files, or interpret result outputs	sub-skills/evaluation/SKILL.md

Operating boundaries

• Do not run sam run, sam task, sam-rest-cli, plugin installs/builds, catalog browsers, broker containers, LLM calls, or sam eval unless the user asked for live execution and the needed services/credentials are available.
• Do not rely on the original repository checkout for operating instructions. This skill's references and scripts are self-contained.
• If a task requires editing this repository rather than using SAM, treat that as repo maintenance and select normal coding-agent workflows; this skill is for operating the package.
• If package version, command names, public config schema, or evidence paths differ from references/repo-provenance.md, refresh the skill before relying on detailed instructions.