Userback platform help — visual customer-feedback and bug-reporting: in-app widget with annotated screenshots, video, auto-captured console/network logs, and session replay, plus feature-request boards, public roadmap, surveys, and a Feature Portal. Developer surface: REST API (rest.userback.io/1.0, Bearer), webhooks, a ~12-tool MCP server (OAuth) for Claude/Cursor/ChatGPT, and JS/React/Vue/Next/mobile SDKs. Use when setting up the Userback widget, identifying logged-in users in feedback, syncing feedback to Jira/Linear/a CRM via API or webhooks, pulling feedback or session recordings programmatically, configuring the Userback MCP server, hitting API/webhook plan gates (Business Plus), or comparing Userback plan tiers. Do NOT use for choosing among feedback tools or NPS/CSAT survey methodology (use /sales-customer-feedback).
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Userback platform help — visual customer-feedback and bug-reporting: in-app widget with annotated screenshots, video, auto-captured console/network logs, and session replay, plus feature-request boards, public roadmap, surveys, and a Feature Portal. Developer surface: REST API (rest.userback.io/1.0, Bearer), webhooks, a ~12-tool MCP server (OAuth) for Claude/Cursor/ChatGPT, and JS/React/Vue/Next/mobile SDKs. Use when setting up the Userback widget, identifying logged-in users in feedback, syncing feedback to Jira/Linear/a CRM via API or webhooks, pulling feedback or session recordings programmatically, configuring the Userback MCP server, hitting API/webhook plan gates (Business Plus), or comparing Userback plan tiers. Do NOT use for choosing among feedback tools or NPS/CSAT survey methodology (use /sales-customer-feedback).
If references/learnings.md exists, read it first for accumulated platform knowledge.
What are you trying to do?
A) Install / configure the feedback widget (annotated screenshots, video, console + network capture) and identify the logged-in user via the JS SDK
B) Read feedback/comments/projects/session recordings out via the REST API (rest.userback.io/1.0, Authorization: Bearer)
C) Push feedback in, or sync to Jira / Linear / GitHub / Slack / a CRM via native connectors, webhooks, or Zapier
D) Set up webhooks (feedback + comment events) or the MCP server for Claude/Cursor/ChatGPT
E) Use feature-request boards + roadmap + Feature Portal, session replay, or surveys
F) Pick a plan / hit a plan gate (API & webhooks are Business Plus; session replay + SDK = Business), or compare Userback vs Marker.io/BugHerd/Usersnap/Canny
Which plan are you on? This gates everything programmatic. API + webhooks + mobile SDK require Business Plus; session replay, surveys, and JS-SDK methods require Business; Zapier requires Team. Free retains feedback only 7 days.
Skip-ahead rule: if the user's prompt already contains enough context, skip to Step 2.
Step 2 — Route or answer directly
If the question is about...
Route to...
Choosing a feedback/bug tool across the market, or VoC / NPS/CSAT/CES survey-program strategy
/sales-customer-feedback {question}
Aggregating unsolicited feedback across reviews/social/support
/sales-noisely {question}
Public review generation strategy (Trustpilot/G2)
/sales-customer-reviews {question}
In-app message / onboarding survey strategy across tools
/sales-in-app-messaging {question}
Connecting Userback to another tool generically (iPaaS)
/sales-integration {question}
When routing, give the exact command, e.g. "This is a tool-selection question — run: /sales-customer-feedback compare Userback vs Marker.io for visual bug reporting".
Step 3 — Userback platform reference
Read references/platform-guide.md for the full reference — the module map (widget, session replay, surveys, boards/roadmap/Feature Portal, Workflows, Insights) tagged API-/webhook-/UI-only, pricing tiers + plan gates (Free/Team/Business/Business Plus), the data model (Feedback, Project, Workflow, Comment, SessionRecording with JSON shapes), and quick-start recipes (install + identify; paginated export; webhook listener; MCP triage).
Read references/userback-api-reference.md for the developer surface — base URL https://rest.userback.io/1.0, authorization: Bearer <TOKEN> (token from Workspace Settings → API Token), the endpoint inventory (Feedback, Comments, Screenshots, Members, Projects, Session Recordings, Workflows), page/limit pagination (max 50) + OData filter/sort, webhook payload shape, MCP setup, and a Gaps section (rate limits + webhook signature not documented).
Answer using only the relevant section. Don't dump the full reference.
Step 4 — Actionable guidance
Focus on the user's specific situation:
The API is the top tier. REST API and webhooks are Business Plus only; Zapier is Team+. If they're prototyping a CRM/warehouse sync on Free/Team/Business, the API won't be available — budget for Business Plus before building. This is the #1 surprise.
Identify users before they submit. Anonymous feedback can't be tied to an account. Call Userback.identify(id, {name,email,plan}) (and setData({...}) for custom context) in the widget init — these SDK methods require Business.
It's a visual feedback tool first. Userback's edge over board-only tools (Canny/Frill) is the auto-captured console logs, network requests, device metadata, and session replay that make a bug reproducible. Lead with that when the use case is bug reporting; if they only want feature voting, a cheaper board may fit.
Webhooks are unsigned. No documented HMAC. Verify User-Agent: Userback-Webhook, keep the URL secret, dedupe on data.id, and re-fetch via GET /feedback/{id} rather than trusting the thin payload.
MCP for agent-driven triage.claude mcp add Userback https://mcp.userback.io/v1/mcp/ -t http -s user (OAuth). ~12 tools — an agent can search/cluster feedback and read console/network logs without ID-copying.
Mind retention + paging. Free purges feedback after 7 days; list endpoints cap at 50/page; handle 429 with backoff.
If you discover a gotcha, workaround, or tip not covered in references/learnings.md, append it there.
Gotchas
Best-effort from research (2026-06) — modules/pricing/API captured from docs.userback.io, support.userback.io, and userback.io/pricing on this date. Re-verify plan gates, the API surface, and webhook details against current docs/in-account before relying on them.
REST API + webhooks are Business Plus. The biggest integration footgun — they're gated to the top tier ($23–29/seat). Session replay, surveys, and JS-SDK methods are Business; Zapier is Team. Don't promise an API/webhook build on a lower plan.
Webhooks have no documented signature. Payloads are thin and unsigned (User-Agent: Userback-Webhook only). Treat them as untrusted: secret URL, dedupe on id, and re-read via the REST API for authoritative data.
Free retention is 7 days. Feedback older than a week is purged on Free Forever — fine for a trial, not for anything you need to keep. Upgrade (or export) before relying on history.
Steep learning curve / feature sprawl. A recurring review complaint is that Userback is "too complex for simple bug tracking" and onboarding/notification settings overwhelm new users. For a dead-simple board, a lighter tool may fit better — say so rather than overselling.
List paging caps at 50.limit max is 50 records/page; exact rate limits are unpublished. Paginate with page/limit and back off on 429.
Anonymous-by-default. If you don't call identify() before submission, feedback isn't attributed to the user — a common "why are all my reports anonymous?" gotcha.
Plan-gated capture, not just API. Session replay capture and mobile SDK are gated (Business / Business Plus) — confirm the plan unlocks the capture feature, not just the API to read it.
Related skills
/sales-customer-feedback — Feedback/VoC strategy across tools — board-tool selection, survey-program design, metric choice (Userback is one visual-feedback option). Install: npx skills add sales-skills/sales --skill sales-customer-feedback -a claude-code
/sales-userjot — AI-first feedback board with a REST API + MCP server and no per-seat pricing — the contrast when you want feature voting + an agent loop without per-user billing (and don't need visual bug capture). Install: npx skills add sales-skills/sales --skill sales-userjot -a claude-code
/sales-frill — Affordable standalone board with a documented REST API + HMAC-SHA256-signed webhooks — the pick when you need signed webhooks (Userback's are unsigned) and a board over visual bug reports. Install: npx skills add sales-skills/sales --skill sales-frill -a claude-code
/sales-sleekplan — AI feedback suite (auto-categorization + MCP server) with a REST API + a free Indie tier — contrast for AI triage plus an API without Userback's Business-Plus gate. Install: npx skills add sales-skills/sales --skill sales-sleekplan -a claude-code
/sales-fourfour — AI voice-of-customer copilot (OData API + MCP server) that mines existing calls/tickets rather than soliciting widget feedback — pair when you already have rich conversation data. Install: npx skills add sales-skills/sales --skill sales-fourfour -a claude-code
/sales-do — Not sure which skill to use? The router matches any sales objective to the right skill. Install: npx skills add sales-skills/sales --skill sales-do -a claude-code
Examples
Example 1: "Export all our Userback bugs into Snowflake nightly" (developer/automation)
User says: "I want a nightly job that pulls every bug report out of Userback into our warehouse with the screenshot and session-replay links."
Skill does: Confirms the REST API requires Business Plus, then gives the pattern: GET https://rest.userback.io/1.0/feedback with authorization: Bearer <TOKEN> (token from Workspace Settings → API Token), filter=feedbackType eq 'Bug', sort=modified,desc, paginating with page/limit=50 (the max) and backing off on 429. Each Feedback object carries screenshots/session data; for replay URLs, join against GET /sessionRecording. Recommends storing modified as a high-water mark for incremental pulls. Points to the paginated Python recipe in references/platform-guide.md.
Result: A correct, plan-aware, incremental export pipeline.
Example 2: "All our feedback comes in anonymous" (troubleshooting / SDK)
User says: "Feedback shows up but with no user — I can't tell who reported what."
Skill does: Explains Userback is anonymous unless you identify the user, and shows the fix: call Userback.identify('<userId>', { name, email, plan }) (plus Userback.setData({...}) for org/MRR context) in the widget init, before submission. Flags that identify/setData and session replay require Business. Notes you can also prefill fields, and that for backfilling attribution you'd update records via the API.
Result: Attributed feedback with the right SDK call and plan caveat.
Example 3: "Is Userback the right tool, or should we use Marker.io / Canny?" (selection)
User says: "Small SaaS team. We want users to report bugs with screenshots and also vote on features. Userback vs Marker.io vs Canny?"
Skill does: Frames the axis: Userback = visual bug capture (annotated screenshots/video + console/network logs + session replay) plus feature boards/roadmap in one tool — but API/webhooks are Business Plus and webhooks are unsigned; Marker.io = deeper two-way issue-tracker (Jira) sync for QA/dev workflows; Canny/Frill/UserJot = stronger pure feature-voting boards (Frill has signed webhooks, UserJot has an MCP server + no per-seat). Recommends Userback when bug-reproduction context matters most and you want voting in the same tool; routes the broader comparison: "run: /sales-customer-feedback compare visual bug tools vs feature-voting boards".
Result: A capability-grounded pick with the right cross-skill routing.
Troubleshooting
The API returns 401 / I can't find where to build the integration
Symptom: API calls fail with 401, or there's no API option in the UI.
Cause: Two things — the REST API is Business Plus only, and the token must be created in Workspace Settings → API Token → Create API Token, then sent as authorization: Bearer <TOKEN>.
Solution: Confirm the workspace is on Business Plus (Team/Business have no API). Generate a token, send it in the authorization header (note: lowercase header, Bearer prefix). Tokens are account-wide; use a separate token per account. If you only have Team, use Zapier instead, or upgrade.
My webhook fires but I can't trust / verify the payload
Symptom: You receive webhook calls but there's no signature to validate, and the body is thin.
Cause: Userback webhooks (Business Plus) send User-Agent: Userback-Webhook and a payload of action / type / data / timestamp / url — no documented HMAC signature.
Solution: Keep the endpoint URL secret, verify the User-Agent, dedupe on data.id + action (ordering/delivery aren't guaranteed), and re-fetch the authoritative record via GET /feedback/{id} (or /feedback/comment/{id}) before acting. Use the built-in test feature under Connect → Webhooks to validate setup.
Old feedback disappeared / reports lack reproduction detail
Symptom: Feedback older than a week is gone, or bug reports don't have enough to reproduce.
Cause: Free Forever retains feedback only 7 days. And session replay + full technical capture are plan-gated (session replay = Business).
Solution: Upgrade past Free (or export via API) to keep history. To get reproducible bugs, ensure you're on Business+ for session replay, and rely on the auto-captured console logs / network requests / device metadata that ship with each report — surface those to your devs (or push them into the linked Jira/Linear issue).
