Webflow Mcp:review Comments

Review open comment threads on a Webflow site and triage each one.

Input: $ARGUMENTS — a site name (e.g. “Workhaus”), site ID (e.g. 6808fd4eff835ee3af009d6f), or either with the -reply flag (e.g. Workhaus -reply).

Flags:

• -reply — in addition to writing the report, post a bot reply to each non-open thread. Without this flag, the skill runs in report-only mode (read-only).

Parse $ARGUMENTS at the start: strip -reply from the input to get the site identifier, and set replyMode = true if -reply was present, false otherwise.

Step 1 — Resolve the site

If the site identifier is empty or blank — cross-site comment survey

1. Call data_sites_tool > list_sites to get all sites. Page through until all are collected.
2. For each site, call data_comments_tool > list_comment_threads with isResolved: false and limit: 100. Page through until all unresolved threads are collected. If the API does not support isResolved filtering, fetch all threads and filter client-side to isResolved === false. Process sites in batches of 20 (batch multiple actions in a single tool call). After each batch completes, log a progress line: Batch {N}/{total} done (sites {start}–{end}): {summary of findings, e.g. "all 0 threads" or "SiteName has X threads, rest 0"}.
3. For each site, compute:
   • unresolvedCount = total unresolved threads
   • newestDate = the maximum lastUpdated value across all unresolved threads for that site (ISO → human-readable date, e.g. “Apr 3, 2026”). If no threads, show —.
   • oldestDate = the minimum lastUpdated value across all unresolved threads for that site (ISO → human-readable date). If no threads, show —.
4. Sort sites by unresolvedCount descending. Take the top 10.
5. Above the table, show a heading line: ### Checked {totalSiteCount} sites — {sitesWithUnresolved} have unresolved comments.
6. Display a table in this format (link just the site name to https://webflow.com/design/{siteId}):

| Site | Unresolved Comments | Newest / Oldest |
|------|---------------------|-----------------|
| [Site Name](https://webflow.com/design/{siteId}) ({siteId}) | {N} | {newestDate} / {oldestDate} |

7. After the table, tell the user: “Run /webflow-mcp:review-comments <site name or ID> to review a specific site.”
8. Write the survey output to a file:
   • Ensure comment-reviews/ exists (create with mkdir comment-reviews if not).
   • Filename: comment-reviews/triage-report-{YYYY-MM-DD-HH-MM}.md using the current local time (zero-padded).
   • File contents: a bold H1 title # **Webflow Comment Review**, then a blank line, then an H3 line ### Report created on: {human-readable date, time, and timezone, e.g. "April 16, 2026 at 10:39 AM PDT"}, then a blank line, then the H3 heading line (### Checked …) and the full table from steps 5–6 above, in markdown.
   • Log the path after writing, e.g. Report written to comment-reviews/triage-report-2026-04-16-14-30.md.
9. Stop — do not proceed to Step 2.

If the site identifier looks like a Webflow site ID (24-char hex), use it directly.

Otherwise call data_sites_tool > list_sites and find the site whose displayName matches the identifier (case-insensitive). If no match, tell the user and stop.

Step 2 — Fetch all open threads

Always make a fresh API call here — never reuse thread data from earlier in the conversation. The user may have added or resolved comments since the last run.

Call data_comments_tool > list_comment_threads with isResolved: false and limit: 100. Page through results until all threads are collected.

Log: Site: {displayName} and Found {N} open thread(s).

Step 3 — Build page-level element map

From the already-fetched thread list, build a frequency map of elementId.element → Set<pageId> across all threads. Any elementId.element that appears on 2 or more distinct pages is almost certainly the page root/body element (a real element ID would be page-scoped; only shared structural roots repeat across pages).

No API calls needed — this is a local computation on the thread data.

Step 4 — Triage each thread

For each thread:

4a — Fetch replies

Call data_comments_tool > list_comment_replies for this thread.

4b — Dedup check

Look for replies whose content includes the string — 🤖 Comment Review Agent.

If found, note the most recent one (lastAgentReply). If no human reply exists with a createdOn after lastAgentReply.createdOn, skip this thread (increment skipped count, continue to next thread).

4c — Compute element context

• elementId = thread.elementId?.element
• isPageLevel = elementId appears on 2 or more distinct pageIds in the frequency map from Step 3

4d — Classify the thread

Use the following criteria:

noise — No real design or engineering value:

• Test/placeholder text (“hello world”, “testing”, “asdf”, random characters)
• Casual reactions with no ask (“looks nice”, “nice!”, “hey hey”)
• Duplicate sentiments that add nothing

stale — Real concern, but old and likely handled:

• Substantive comments older than 14 days with no replies and no follow-up
• Questions that are probably resolved (“beta for how long?”, “is this good contrast?”)
• Action items that normal review would have caught

open — Real, actionable concern needing attention:

• Explicit tasks (“Should be sentence case”, “fix image”, “Look at name wrt L10N”)
• Design decisions still required
• Specific and concrete concerns

page-level — Comment is on the page root, not a specific element:

• Use this when isPageLevel is true
• The comment is not anchored to any specific element — it may be intentional or may be an orphan from a deleted element

4e — Compose reply

Always compose the reply text (it appears in the report regardless of mode):

• noise: one sentence confirming it’s safe to resolve. E.g. "Looks like test text — safe to resolve."
• stale: state the age in days, suggest resolving, invite reopen. E.g. "This is 302 days old with no follow-up. Safe to resolve — reply here if it's still relevant."
• open: no reply text — surface in report only.
• page-level: one sentence noting it’s not attached to a specific element. E.g. "This comment is on the page root rather than a specific element — it may be an orphan from a deleted element. Safe to resolve if no longer relevant."

Append \n\n— 🤖 Comment Review Agent to every reply text.

4f — Post reply (only if replyMode = true)

If replyMode is true and verdict is not open, call data_comments_tool > create_reply with the composed reply content.

Log each thread as:

{VERDICT_EMOJI} {verdict}  "{preview (60 chars)}"
                {one-sentence classification reason}
                ↳ {reply posted | no reply — report only | skipped}

Verdict emojis: noise = 🗑, stale = 🕰, open = 🔴, page-level = 📄

Step 5 — Write the report

After processing all threads:

1. Check whether a comment-reviews/ directory exists at the top level of the working directory. If it does not exist, create it with mkdir comment-reviews.
2. Write a markdown report to comment-reviews/{slugified-site-name}-comments-triage-report.md (lowercase, hyphens, no special chars).

Report format:

# Comment Review — [{siteName} ({siteId})](https://webflow.com/design/{siteId})

**Run:** {human-readable date and time}
**Mode:** {Report only | Report + replies posted}
**Threads:** {total} total | 🔴 {open} open | 🕰 {stale} stale | 📄 {page-level} page-level | 🗑 {noise} noise | ⏭ {skipped} skipped

---

Use _None._ for any section with no entries.

Log a summary line and confirm the report path.