clip-thread

name	clip-thread
description	Capture a chat or email thread from a connected MCP to the knowledge intake. Use when user says "/aria-cowork:clip-thread" or "archive this conversation". Pulls thread content from ~~chat (Slack, Teams) or ~~email (Gmail, MS365) MCP into intake/clippings/ for review. (Cowork variant — namespaced-only.)
argument-hint	<thread-url-or-id> [tags]

/clip-thread — Capture Chat/Email Thread to Intake

Save a chat thread or email conversation to intake/clippings/{YYYY-MM-DD}-{slug}.md for review and promotion. Unlike /clip (URL/snippet) or /intake (bulk/doc-anchored), /clip-thread is shaped specifically for threaded conversations — Slack threads, Teams channel discussions, Gmail conversation chains, MS365 email threads.

Runtime Gate (per ADR-094)

Canonical resolution: This is the Claude Cowork variant — namespaced-only. When both plugin-claude-code and plugin-claude-cowork are loaded in the same session (most common in Claude Desktop), bare /clip-thread resolves to aria-knowledge's variant — Code is the canonical owner of all 24 dual-port skills per ADR-094 §Part 1. To reach this skill, use the namespaced form: /aria-cowork:clip-thread. Do NOT match bare /clip-thread — that belongs to aria-knowledge.

Before Step 0: Check whether the Bash tool is available in this session. If Bash IS available (you are running in Claude Code or another runtime with shell access), surface the following notification and wait for explicit user confirmation:

⚠️ Runtime mismatch — you invoked aria-cowork's /clip-thread from a runtime with shell access.

This skill is Cowork-native because it depends on ~~chat / ~~email MCPs that are typically only connected in Cowork. If you're in Code and DO have these MCPs configured there, fine to proceed; otherwise the canonical aria-knowledge variant at /clip-thread will at least surface the same gate.

Use /clip-thread instead? (y / n)

Wait for an explicit reply:

• y / yes — Use the Skill tool to invoke clip-thread (the bare-slash canonical, which routes to aria-knowledge when both ports are loaded) with the same arguments the user provided to this invocation. Do not proceed with this skill's steps; the aria-knowledge variant takes over and runs to completion. This is the default-yes path — auto-redirect is the helpful action.
• n / no — Proceed with this (aria-cowork) variant anyway despite the runtime mismatch. The user has explicitly opted in.
• No response / any other reply — Treat as "do not proceed" and exit cleanly without running either variant.

This gate applies even when mode = auto per ADR-094 §Part 3. Auto mode's "implicit-yes on all gates" rule is suspended for the runtime-mismatch check — auto trusts that the user invoked the correct variant, and this gate enforces that precondition. All other auto-mode gates remain bypassed. The friction cost is now low: on y, the auto-redirect runs the correct variant with the original args.

If Bash is NOT available (normal Cowork runtime), proceed to Step 0.

Step 0: Resolve config

The default knowledge folder is ~/Projects/knowledge/ (expand ~ to your home directory's absolute path). Read <knowledge_folder>/aria-config.md and extract knowledge_folder (the absolute path).

If aria-config.md doesn't exist, stop with: "aria-cowork is not configured. Run /aria-setup to get started."

Verify <knowledge_folder>/intake/clippings/ exists. If not, stop with: "Clippings directory not found. Run /aria-setup to repair the knowledge folder structure."

For all subsequent file operations in this skill, use the absolute path from knowledge_folder directly. Cowork resolves absolute paths via the persistent grant from claude_desktop_config.json per ADR-008. aria-knowledge in Code uses the same absolute path to reach the same files.

Step 1: Probe Connected MCPs

Check Claude's available tool list for ~~chat and ~~email MCPs. The skill needs at least ONE of these connected to proceed.

• ~~chat (slack, ms365): if connected, available for Slack threads or Teams channel messages.
• ~~email (gmail, ms365): if connected, available for Gmail or MS365 email threads.

If NEITHER category has a connected MCP, output the standard fallback notice and stop:

No required MCPs connected for /clip-thread. Connect one of: Slack, Microsoft 365 Teams (for ~~chat) or Gmail, Microsoft 365 Outlook (for ~~email) via Cowork Settings → Connectors (or Claude Code's .mcp.json for the Code surface). See CONNECTORS.md. Skipping this run.

Per ADR-015, the runtime tool list IS the probe — no explicit API call needed.

Step 2: Parse Input

The user provides one of:

• A Slack thread URL — https://<workspace>.slack.com/archives/<channel>/<message-ts> or thread permalink
• A Teams message link — https://teams.microsoft.com/l/message/...
• A Gmail thread ID or URL — https://mail.google.com/mail/u/0/#inbox/<thread-id> or just the thread ID
• An MS365 message ID — opaque ID from Outlook
• A bare thread/conversation identifier — opaque string the connected MCP can resolve

Optionally followed by tags (e.g., slack #engineering, email customer-feedback).

Source-type detection:

Input shape	Routes to
Contains slack.com/archives/	~~chat (slack)
Contains teams.microsoft.com	~~chat (ms365)
Contains mail.google.com or matches Gmail thread-ID pattern	~~email (gmail)
Outlook-shaped ID	~~email (ms365)
Ambiguous	Ask user which category to route through

If the detected category's MCP is NOT connected, output the category-specific fallback notice and stop:

This thread looks like a [Slack thread / Gmail thread / etc.] but the ~~chat / ~~email MCP for that source is not connected. Connect [vendor] via Cowork Settings → Connectors and retry. Skipping this run.

Step 3: Fetch Thread Content

Call the appropriate MCP tool to retrieve the thread:

• Slack: use the connected slack MCP's conversations.replies (or equivalent) tool with the channel + thread_ts derived from the URL. Retrieve all messages in the thread.
• Teams: use the connected ms365 MCP's message-thread tool.
• Gmail: use the connected gmail MCP's messages.get for each message in the thread (or threads.get if exposed).
• MS365 Outlook: use the ms365 MCP's email-thread tool.

For each message in the thread, gather:

• Author (display name + handle/email)
• Timestamp (ISO format)
• Body (plaintext if available, markdown if the source supports it)
• Reactions / attachments (if exposed — note their presence but don't fetch attachment bodies in this skill)

Limit: if the thread has more than 50 messages, fetch the first 50 + note the truncation in the clipping. The user can run /clip-thread again with a refined identifier for the tail.

Step 4: Compose Clipping

Slug-ify the thread for the filename: use the first message's first ~6 significant words, lowercased, hyphen-separated, ASCII-only.

Generate the date stamp as YYYY-MM-DD (local time).

Filename: intake/clippings/{date}-{slug}.md

If a file with that name already exists, append -2, -3, etc. to deduplicate.

Body template:

---
source: <thread-url-or-id>
source_type: <slack|teams|gmail|ms365>
date: <YYYY-MM-DD>
participants: [<list of unique authors>]
message_count: <N>
tags: [<from input args, plus auto-detected like 'slack' or 'email'>]
---

# <thread topic — extracted from first message's first line or the channel name>

## Context

- **Source:** <source URL or "Slack thread in #channel" / "Gmail thread with N replies">
- **Participants:** <N authors>: <list>
- **Spans:** <first-message-date> → <last-message-date>
- **Message count:** <N> (or "<N> of <total> — truncated at 50")

## Thread

<for each message:>

### <Author> — <timestamp>

<message body, lightly cleaned (remove Slack/MS markup if obvious, preserve markdown)>

<reactions if present: e.g., "Reactions: 👍 ×3, 🎉 ×1">

<attachment notes if present: e.g., "[attachment: deploy-logs.txt (45KB)]">

---

## Reaction

<intentionally left empty — the user's reaction / why this is worth keeping. /audit-knowledge surfaces this for review at next audit pass.>

Step 5: Write + Report

Write the composed clipping to <knowledge_folder>/intake/clippings/{date}-{slug}.md.

Report to user:

Clipped <N> messages from <source-type> thread to intake/clippings/<date>-<slug>.md.

- Participants: <list, max 5 + "and N more" if truncated>
- Spans: <date range>
- Tags: <from input>

Next: review at next /audit-knowledge run. Add a reaction in the "## Reaction" section if you want to lock in the why-it-matters context now.

Rules

• Never auto-promote. Clippings stay in intake/clippings/ until /audit-knowledge routes them. This skill is intake-only.
• Never delete the source. Reading a thread via MCP doesn't modify it; this is read-only externally.
• Respect 50-message truncation cap. Do NOT loop and fetch all pages. If user wants the full thread, they can re-invoke with a refined scope or use the source platform's native export.
• Strip secrets if obvious — if the thread contains what look like API keys, OAuth tokens, or passwords (regex match against common patterns), replace with [REDACTED] in the body and add a redactions: true field to frontmatter. Surface in the report.
• One thread per invocation. If the user wants multiple threads, suggest /intake bulk mode.

Notes

• Composes with /audit-knowledge — clippings written here are routed at next audit per the standard intake-disposition vocabulary (Accept → tracker / roadmap / todo / adr / plan / bundle / rule, or Reject / Defer).
• The ## Reaction section is the user's voice slot. /clip-thread never fills it (the same precedent /intake doc set in v2.17.0 / v0.3.0).
• This skill is bidirectional per ADR-014 — aria-knowledge v2.18.0 ships an identical body. Output schema is byte-identical per ADR-013.
• Probe semantics per ADR-015.