Analyze customer recommendation and satisfaction surveys, NPS and CSAT, from Google Sheets. Group comments into themes, explain weekly score changes and low-score drivers, and post a Slack digest with representative quotes without modifying survey data.
中文简介正在准备中。当前显示英语简介。
Weekly NPS/CSAT theme analysis for {{survey_sheet}}. Reads survey responses from Google Sheets, clusters them into themes with sentiment and detractor drivers, tracks score movement week over week, and posts a summary with representative quotes to {{report_channel}}. Read-only on the sheet and report-only — no writes to any survey system, no respondent contact.
安装前可以查看交给 AI 代理的指令内容。
Read-only and report-only: the agent never edits the sheet and never contacts a respondent. The output is a Slack message; a human decides what to act on. </overview>
<when-to-load> - The weekly cron fires the NPS/CSAT analysis sweep. - A human asks the agent to analyze survey responses or summarize NPS/CSAT. </when-to-load> <workflow>Read {{survey_sheet}} in full through the google_sheets connector,
read-only. Identify the columns: response timestamp, score (0–10 for NPS,
typically 1–5 for CSAT), free-text comment, and any segment/plan column if
present. If column names vary, match by content (a 0–10 numeric column next to
free text is almost always the NPS score).
Using the timestamp column, split responses into "this week" (since the last run, one cadence period back) and "prior periods" (everything before). Because the session is fresh each run, this split is derived entirely from the sheet's own dates — there is no external ledger to consult.
For NPS: promoter (9–10), passive (7–8), detractor (0–6). For CSAT: map the top band to promoter-equivalent, middle to passive-equivalent, bottom to detractor-equivalent, using whatever scale the sheet actually uses. For each response with a comment, tag sentiment (positive / neutral / negative) from the text itself — a detractor can still leave a neutral comment, and a promoter can still flag a real complaint.
Group the free-text comments into themes (e.g. pricing, onboarding, support responsiveness, a missing feature, performance, a specific bug) by what the respondent means, not the exact wording, using the full response history so a theme is recognized as recurring even when no two comments phrase it the same way. For each theme, pick one or two representative quotes and note which score band they came from.
Filter to this week's detractor responses specifically. Rank the themes that show up among detractors by frequency, and call out the top few as "detractor drivers" with a count and a quote each. Compare against prior weeks' detractor themes: flag any driver that is new this week or has grown in share.
Compute this week's NPS/CSAT score (using the standard formula: %promoters − %detractors for NPS, or the equivalent top-box rate for CSAT) from the full history and compare it to the prior period(s) in the same sheet. State the direction and size of the move, and attribute it: is it more detractors, fewer promoters, a shift in passives, or some combination.
Post one message to {{report_channel}}:
Nothing else is written anywhere. The sheet is never modified, and no respondent is contacted.
</workflow> <guardrails> - **Read-only on the sheet.** The agent never edits a row, adds a column, or writes a score back to `{{survey_sheet}}`. - **Report only, no action.** The agent never emails, messages, or otherwise contacts a survey respondent, and never opens a ticket or task on their behalf. Deciding what to do about a theme stays with people. - **Single output.** The only thing that leaves the sandbox is the one weekly post to `{{report_channel}}`. - **Secrets scoped.** The Google Sheets and Slack credentials are injected at runtime; never written to disk or logged. - **Recompute, don't accumulate.** Each run is a fresh session — themes, bands, and the trend are always recomputed from the sheet's own history, not from a memory file, so there's nothing stale to carry forward or drift from. </guardrails> </skill>还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Review Slack access requests against Okta role and team information, then prepare narrowly scoped GitHub or AWS IAM permissions. Apply changes only after an eligible person's approval and record the grant, reasoning, approver, and timestamp.
中文简介正在准备中。当前显示英语简介。
Research companies and people before outreach or meetings. Combines business profiles, recent news, hiring signals, and key contacts into a sourced brief with talking points and suggested questions, adding relationship history when CRM data is connected.
中文简介正在准备中。当前显示英语简介。
Review daily Google Ads and Meta Ads results for budget pacing, declining performance, and unusual activity. Get prioritized findings with supporting numbers and concrete recommendations for a marketing team to assess and carry out.
中文简介正在准备中。当前显示英语简介。
Automate website tasks such as navigating pages, filling forms, clicking buttons, taking screenshots, and extracting data. The skill also supports web app testing, with specialized guides for Electron desktop apps, Slack, and cloud browsers.
中文简介正在准备中。当前显示英语简介。
Process emailed vendor invoices by extracting details, comparing them with purchase orders and invoice history, and recording duplicates, overcharges, or missing orders. Summarize results in Slack for human review; payment approval and execution remain with people.
中文简介正在准备中。当前显示英语简介。
Guide changes to Kortix billing code, admin controls, and pricing screens. Explain how plans, trials, enterprise contracts, and access exceptions should work, where pricing is defined, and which audited routes must handle account changes.
中文简介正在准备中。当前显示英语简介。