Agent Skills › overmind-core/overmind › overmind-integrations

overmind-integrations

GitHub

用于连接外部追踪提供商到 Overmind,检查连接器健康状态,审查能力边界映射并验证导入的追踪数据。

overmind/skills/overmind-integrations/SKILL.md overmind-core/overmind

Trigger Scenarios

连接外部追踪服务 配置追踪映射边界 验证追踪数据同步

Install

npx skills add overmind-core/overmind --skill overmind-integrations -g -y
More Options

Non-standard path

npx skills add https://github.com/overmind-core/overmind/tree/main/overmind/skills/overmind-integrations -g -y

Use without installing

npx skills use overmind-core/overmind@overmind-integrations

指定 Agent (Claude Code)

npx skills add overmind-core/overmind --skill overmind-integrations -a claude-code -g -y

安装 repo 全部 skill

npx skills add overmind-core/overmind --all -g -y

预览 repo 内 skill

npx skills add overmind-core/overmind --list

SKILL.md

Frontmatter
{
    "name": "overmind-integrations",
    "description": "Connect external tracing providers to Overmind, inspect connector health, review capability-boundary mappings and verify imported traces. Use for the Integrations surface; repository SDK instrumentation belongs to Observability."
}

Overmind Integrations

Start with list_projects and choose the intended accessible project. For an account connection, pass its project_id on every project tool and resource URI query; follow returned links. Project API keys retain their narrower access.

Bring external traces into the correct project with explicit boundary mapping. Use the configured Overmind MCP connection and read overmind://project/current?project_id=ID. The provider's source project and the Overmind destination are different identities.

Inspect and authenticate

Start with inspect_connectors. For a supported provider, prefer the native connect-traces prompt. Use the server's available provider list and setup commands rather than assuming that a named service is supported.

Credentials are not MCP arguments. If inspection returns connector_setup_required, show its CLI command and let the human enter the provider credential in their terminal. Do not ask for a key in chat, export it, or run an interactive credential flow in a non-TTY sandbox. Read overmind://connector-setup for the current handoff. Continue after the human confirms completion or supplies the non-secret connector ID.

Review the mapping

Inspect the connector with include_source_projects=true. Resolve the chosen source project and lookback. Examine observation_shapes, suggested_boundaries, alternatives, unmapped_roots and mapping_options.

mapping.names declares capability boundaries that become Overmind trace roots. Default to the suggested parent observation names so children remain nested. An alternative can replace that parent when chosen; do not include both it and its ancestor. Tool/LLM spans and other nested names are not independent roots unless the human chooses that boundary.

Use configure_connector to propose the selected source, lookback and mapping without confirm_mapping. Present the concrete mapping and the import-unmapped option, then wait for its review. After approval, confirm that same mapping with confirm_mapping=true. Do not silently broaden the import window or destination.

Sync and verify

Run the authorized sync_connector, poll with get_job(kind=connector_sync, id=...), then query the imported traces. Report imported/skipped counts, failed work and remaining unmapped roots from the actual result. A stored credential or confirmed mapping is not proof that traces arrived.

For inspection-only work, read overmind://connectors/{connector} and status without starting another sync. There are no MCP deletion or credential-rotation tools to invent.

Return the source/destination identities, mapping decision, sync outcome and console_traces_url. For configuration, open observability/integrations under the project's Console base with its projectId.

Version History

  • e146ecd Current 2026-10-18 15:42

Same Skill Collection

.agents/skills/api-endpoints/SKILL.md
.agents/skills/backend-architecture/SKILL.md
.agents/skills/code-comments/SKILL.md
.agents/skills/data-workshop/SKILL.md
.agents/skills/engineering-taste/SKILL.md
.agents/skills/frontend-design/SKILL.md
.agents/skills/mcp/SKILL.md
.agents/skills/pr-etiquette/SKILL.md
.agents/skills/run-tests/SKILL.md
.agents/skills/seed-demo-data/SKILL.md
overmind/skills/model-training-engineering/SKILL.md
overmind/skills/overmind-agent/SKILL.md
overmind/skills/overmind-datasets/SKILL.md
overmind/skills/overmind-evaluations/SKILL.md
overmind/skills/overmind-inference/SKILL.md
overmind/skills/overmind-observability/SKILL.md
overmind/skills/overmind-optimiser/SKILL.md
overmind/skills/overmind-training/SKILL.md
overmind/skills/overmind/SKILL.md
.agents/skills/finetuning-model-onboarding/SKILL.md

Metadata

Files
0
Version
e146ecd
Hash
ad748e02
Indexed
2026-10-18 15:42

홈 - 위키
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-10-19 05:42
浙ICP备14020137号-1