mcpbeat

Channel Forward Parity

vercel-labs/channel-forward-parity

Webhook route parity audit for channel delivery changes: ensure terminal paths log, record lastForward, classify failures, and refresh stale sandbox port URLs.

709 tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
117
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/vercel-labs/vercel-openclaw --skill channel-forward-parity

The instruction itself

5 sections, as written by the author

Channel Forward Parity

Use this before modifying any channel webhook route or the shared drain-channel workflow.

Source Files To Audit

  • src/server/channels/last-forward.ts
  • src/server/workflows/channels/drain-channel-workflow.ts
  • src/app/api/channels/slack/webhook/route.ts
  • src/app/api/channels/telegram/webhook/route.ts
  • src/app/api/channels/discord/webhook/route.ts
  • src/app/api/channels/whatsapp/webhook/route.ts
  • src/app/api/channels/summary/route.ts
  • src/server/admin/why-not-ready.ts

Required Output

Produce a table:

| Channel | Branch | Logs event | Updates lastForward | Classification | Stale URL refresh | Evidence |

|---|---|---|---|---|---|---|

Also include correlation coverage:

| Channel | Branch | requestId/deliveryId propagated | workflowRunId propagated | Logs event | Updates lastForward | Classification | Stale URL refresh | Evidence |

|---|---|---|---|---|---|---|---|---|

Rules

  • Every delivery attempt must update lastForward.
  • Every skip/reject branch must log a structured reason.
  • Workflow forwarding already records lastForward; do not double-record above it.
  • Treat route-ready, native-accepted, and user-visible-reply separately.
  • Do not add info-level logs on hot polling paths unless operationally necessary; ring-buffer eviction hides evidence.
  • Never add export const runtime = "nodejs" to route handlers.
  • For workflow handoff branches, verify that channels.<ch>_workflow_started, channels.forward_attempt, and channels.forward_outcome can be correlated by requestId/deliveryId and workflowRunId when available.
  • If npx workflow cannot inspect the run, parity evidence must fall back to Vercel logs plus /api/admin/logs.
  • Do not accept "workflow started" as proof of native acceptance; require forward attempt/outcome or user-visible reply evidence.
  • Do not accept npx sandbox failure as sandbox-unreachable until the app admin SSH/exec fallback has been attempted or marked unavailable.
  • Verify project targeting before trusting vercel, npx workflow, or npx sandbox output. If .vercel/project.json points at a different release, require explicit project/team/deployment targeting evidence.

Current Audit Seeds

  • Check Slack fast-path fetch-exception branches for recordChannelLastForward coverage.
  • Check Telegram unauthorized, invalid JSON, and dedup branches for structured logs where operator evidence is needed.
  • Check WhatsApp duplicate/skip branches for structured logs and clear classifications.
  • Check every sandbox-not-listening branch for exactly one stale-port refresh per request.

How to use it

Copy the folder

Take vercel-labs/channel-forward-parity from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.

Install what it needs

The instructions reference npx. Without those the skill loads but fails at the first command.