AGENT SKILL

pm-vercel

Vercel connector knowledge and safe action guide.

License 许可
AGPL-3.0
所属仓库
polymetrics-ai/cli
最近更新
2026-09-21
Stars
★ 7
Forks
0

01 前言 Front Matter

name
pm-vercel
description
Vercel connector knowledge and safe action guide.

03 SKILL.md 正文

pm-vercel

Purpose

Reads deployments, projects, teams, domains, aliases, webhooks, log drains, and edge configs from the Vercel REST API, and writes projects, deployments, domains, project environment variables, webhooks, log drains, edge configs, and alias removal.

Icon

  • id: simple-icons-vercel
  • asset: icons/simple-icons/vercel.svg
  • title: Vercel
  • simple_icon_slug: vercel
  • simple_icon_hex: 000000
  • source: simple-icons
  • license: CC0-1.0
  • review_status: cc0_with_trademark_caveat
  • review_url: https://simpleicons.org/?q=Vercel
  • match: exact-name-or-slug
  • matched_by: vercel

Capabilities

  • check=true catalog=true read=true write=true query=false
  • Integration type: api

Authentication

  • Use pm credentials add with --from-env or --value-stdin for secret fields.

Configuration

  • base_url
  • start_date
  • team_id
  • access_token (secret) (required)

ETL Streams

  • deployments:
    • primary key: id
    • cursor: created
    • fields: created(integer), id(string), name(string), state(string)
  • projects:
    • primary key: id
    • fields: accountId(string), createdAt(integer), framework(string), id(string), name(string), updatedAt(integer)
  • teams:
    • primary key: id
    • fields: id(string), name(string), slug(string)
  • domains:
    • primary key: id
    • fields: createdAt(integer), id(string), name(string), teamId(string), verified(boolean)
  • project_env_vars:
    • primary key: id
    • fields: createdAt(integer), id(string), key(string), project_id(string), target(array), type(string), updatedAt(integer)
  • aliases:
    • primary key: uid
    • fields: alias(string), created(string), createdAt(integer), deployment(object), deploymentId(string), projectId(string), uid(string)
  • webhooks:
    • primary key: id
    • fields: createdAt(integer), events(array), id(string), ownerId(string), projectIds(array), updatedAt(integer), url(string)
  • log_drains:
    • primary key: id
    • fields: createdAt(integer), deliveryFormat(string), environments(array), id(string), name(string), projectIds(array), samplingRate(number), sources(array), updatedAt(integer), url(string)
  • edge_configs:
    • primary key: id
    • fields: createdAt(integer), digest(string), id(string), itemCount(integer), ownerId(string), sizeInBytes(integer), slug(string), updatedAt(integer)

Sync Modes

  • ETL sync modes: full_refresh_append, full_refresh_overwrite, full_refresh_overwrite_deduped

Reverse ETL Actions

  • create_project:
    • endpoint: POST /v11/projects
    • required fields: name
    • risk: external mutation; approval required
  • update_project:
    • endpoint: PATCH /v9/projects/{{ record.id }}
    • required fields: id
    • risk: external mutation; approval required
  • delete_project:
    • endpoint: DELETE /v9/projects/{{ record.id }}
    • required fields: id
    • risk: destructive external mutation; approval required
  • create_deployment:
    • endpoint: POST /v13/deployments
    • required fields: name
    • risk: external mutation; approval required
  • cancel_deployment:
    • endpoint: PATCH /v12/deployments/{{ record.id }}/cancel
    • required fields: id
    • risk: external mutation; approval required
  • delete_deployment:
    • endpoint: DELETE /v13/deployments/{{ record.id }}
    • required fields: id
    • risk: destructive external mutation; approval required
  • add_project_domain:
    • endpoint: POST /v10/projects/{{ record.project_id }}/domains
    • required fields: project_id, name
    • risk: external mutation; approval required
  • remove_project_domain:
    • endpoint: DELETE /v9/projects/{{ record.project_id }}/domains/{{ record.domain }}
    • required fields: project_id, domain
    • risk: destructive external mutation; approval required
  • create_project_env_var:
    • endpoint: POST /v10/projects/{{ record.project_id }}/env
    • required fields: project_id, key, value, type
    • risk: external mutation; approval required
  • delete_project_env_var:
    • endpoint: DELETE /v9/projects/{{ record.project_id }}/env/{{ record.id }}
    • required fields: project_id, id
    • risk: destructive external mutation; approval required
  • create_webhook:
    • endpoint: POST /v1/webhooks
    • required fields: url, events
    • risk: external mutation; approval required
  • delete_webhook:
    • endpoint: DELETE /v1/webhooks/{{ record.id }}
    • required fields: id
    • risk: destructive external mutation; approval required
  • create_log_drain:
    • endpoint: POST /v1/log-drains
    • required fields: deliveryFormat, url, sources
    • risk: external mutation; approval required
  • delete_log_drain:
    • endpoint: DELETE /v1/log-drains/{{ record.id }}
    • required fields: id
    • risk: destructive external mutation; approval required
  • create_edge_config:
    • endpoint: POST /v1/edge-config
    • required fields: slug
    • risk: external mutation; approval required
  • update_edge_config:
    • endpoint: PUT /v1/edge-config/{{ record.id }}
    • required fields: id, slug
    • risk: external mutation; approval required
  • delete_edge_config:
    • endpoint: DELETE /v1/edge-config/{{ record.id }}
    • required fields: id
    • risk: destructive external mutation; approval required
  • delete_alias:
    • endpoint: DELETE /v2/aliases/{{ record.uid }}
    • required fields: uid
    • risk: destructive external mutation (removes a deployment alias); approval required

Security

  • read risk: external Vercel API read of deployment, project, team, domain, alias, webhook, log-drain, and edge-config data
  • write risk: external mutation of Vercel projects, deployments, domains, environment variables, webhooks, log drains, edge configs, and aliases; approval required
  • approval: read: none; write: required for every action
  • Never pass secret values in chat, shell arguments, logs, docs, or JSON output.

Commands

Inspect as a manual

pm connectors inspect vercel

Inspect as structured JSON

pm connectors inspect vercel --json

Agent Rules

  • Run pm connectors inspect vercel before creating credentials or plans.
  • Use --json only when the caller needs structured output; use the manual for human-readable guidance.
  • Never ask the user to paste secret values into chat.
  • For reverse ETL writes, create a plan, show the preview, wait for explicit approval, then run with the approval token.

04 同仓库更多 Skills