CtrlK
BlogDocsLog inGet started
Tessl Logo

agent-switchboard-ops

以"管理者—执行者"模式运维 agent-switchboard MCP:受管 Claude supervisor 生命周期、跨模型委派(Codex/Claude/Gemini)、验收纪律与已知陷阱对策。 用于存在任一可用委派通道(mcp__agent-switchboard__* 工具或宿主原生 subagent),且用户要求委派执行、监督长任务、跨模型审查、回收多 agent 会话成果,或询问"怎么让 Codex/Claude 帮我做"的场景。 不用于单会话即可完成的普通修改;未安装 agent-switchboard 时仍可经原生 subagent 通道触发,但不要猜测工具名。

69

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

77%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with concrete tools, parameters, and validation-gated workflows, but it is padded with incident/citation detail and references example/script files that are absent from the bundle.

Suggestions

Move the dated incident postmortems and external citation URLs into a separate references file (or a 'history/rationale' appendix) and keep SKILL.md to the rules, to improve conciseness.

Provide the referenced bundle files (examples/web-search-via-cli-worker.md, examples/delegate-implementation.md, scripts/evolution_scan.js) or remove the references, so signaled one-level-deep links actually resolve.

Extract the 14-row pitfall table into its own reference document and summarize the top recurring pitfalls inline, reducing inlined bulk while preserving navigability.

DimensionReasoningScore

Conciseness

Informationally dense but padded with dated incident postmortems ('2026-08-24 会话 9 个 supervisor…'), citation URLs, and verbose anti-pattern narration that competes with the context budget; several sections could be tightened.

3 / 5

Actionability

Provides concrete tool names with exact parameters (`route_agent_task(target_model=...)`, `queue_cli_request(backend=codex_cli, target_model=gpt-5.6-luna, effort=medium, ...)`), specific values (`stall_timeout_seconds` 300~600, `API_TIMEOUT_MS=600000`), and copy-paste-ready output-contract templates.

5 / 5

Workflow Clarity

Clear five-step lifecycle (启动→派工→追踪→验收→关闭归档) with explicit validation checkpoints, review→repair feedback loops capped at 2 rounds, and a route-branched verification checklist — appropriate for a destructive/batch-risk skill.

5 / 5

Progressive Disclosure

Has clear section headers, but the large pitfall table and inline citation/anti-pattern bulk that belongs in reference files is inlined, and referenced paths (`examples/web-search-via-cli-worker.md`, `examples/delegate-implementation.md`, `scripts/evolution_scan.js`) do not exist in the bundle.

3 / 5

Total

16

/

20

Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concrete, third-person, and explicitly answers both what and when, with natural trigger phrases and a clear exclusion boundary. It is a strong, well-scoped description.

DimensionReasoningScore

Specificity

Names the domain (agent-switchboard MCP 运维) and multiple concrete actions — 受管 supervisor 生命周期、跨模型委派(Codex/Claude/Gemini)、验收纪律、已知陷阱对策 — covering the lifecycle comprehensively.

5 / 5

Completeness

Explicitly states both what the skill does and when to use it ('用于…且用户要求…' with concrete trigger phrases), plus an explicit when-not clause ('不用于单会话即可完成的普通修改').

5 / 5

Trigger Term Quality

Includes natural user phrases like '怎么让 Codex/Claude 帮我做'、'委派执行'、'监督长任务'、'跨模型审查'、'回收多 agent 会话成果', but a few natural synonyms/variants are missing so it is not fully comprehensive.

4 / 5

Distinctiveness Conflict Risk

A clear niche (agent-switchboard MCP cross-model delegation supervision) with distinct triggers and an explicit exclusion clause, giving minimal conflict risk with other skills.

5 / 5

Total

19

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

15

/

16

Passed

Repository
ooooooooooooooooooop/agent-tools
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.