Use when an interview loop has to be booked, rebuilt, or moved, when a candidate or interviewer needs a message, or when the hiring tracker needs updating or checking for stalls.
Instructions your experts follow when using this skill.
Run this when a loop must be booked, rebuilt, or moved; when a candidate or interviewer needs a message; when the user asks what is stuck; or when they hand over roles and candidates to keep track of.
The tracker; the role and stage; the panel and the competency each person owns; the candidate's availability and timezone; interviewer constraints; and connected calendars or pasted availability. To start or rebuild the tracker, any one source will do: a pasted list, an export from their applicant tracking system (ATS), a spreadsheet or CSV (comma-separated values) file, a link to a shared sheet, or candidate emails they forward.
Agree the number of slots, their lengths, what each covers, and where the breathing room goes. A four-slot onsite gets 10 minutes between interviews plus a proper break near the midpoint. Every window carries two labelled times, the candidate's and the user's; a time with no zone never appears.
Strongest first. Each lists the date, every slot in both timezones, who interviews in each, and the one compromise it makes, such as a tight turnaround or a stand-in interviewer. Once the user picks, hand over three things at once: holds laid out in a copyable table (or drafted invites if Google Calendar is connected); a brief email to the candidate covering times, what each conversation is about, and the whole process in one line (every stage, who they meet, when a decision comes; set this map at the screen stage and keep to it); and a note per interviewer with their slot and competency.
If an interviewer drops out or the loop moves, the competency still needs covering: name a substitute who can take it, or tell the user the slot has no owner yet.
Lead with the answer or the request, state the next step, and give a reply date. The shapes:
Candidates hear back within 24 hours at every stage, 48 at most. If the user has shown you how they write, sound like them. Times come from the tracker or the user, never from you.
These column names are fixed; other skills read them.
Titles, stage names, and notes stay in the user's wording; only dates and times get normalised. Candidate rows match on name plus role, loop rows on candidate plus date; when two copies disagree, the newer wins and you say what changed. Unparseable rows go on a short "needs a look" list. The tracker outranks anything said in chat: read it fresh at the start of each run and save your changes at the end.
Default thresholds, unless the user set their own:
| What | Stalled once |
|---|---|
| Candidate awaiting feedback after an interview | 2 business days |
| Scorecard not filed after the slot | 1 day |
| Offer still open | past its answer-by date |
| Scheduling request with no loop booked | 2 business days |
| Candidate with no reply | 3 touches |
| Interviewer invite not accepted | inside 24 hours of the loop |
| Decline not drafted after the decision | 48 hours |
| Any thread with no update sent | 3 business days |
Group by whoever holds each item, oldest first, one line each with the single action that frees it, plus a drafted nudge. An item stalling a second time gets escalated: who holds it, the days it has cost, and what that does to the loop.
The same read powers the on-demand morning hiring brief (today's interviews with each slot's interviewer, who still needs scheduling by longest wait, and who each open item waits on) and the sweep of candidate and interviewer mail for anything that endangers a booked loop. The sweep flags and drafts; it never replies.
Leave any protected-characteristic column (age or birth date, gender, race, nationality, religion, disability, marital or family status) out of an export and tell the user in one line. Contact details are recorded as given; never guess an email or phone number, and never go hunting for one.
Booking, inviting, cancelling, sending, and declining each wait for the user to approve that one action. Candidate details stay out of group channels.
Ranked loop options, then the holds, candidate email, and panel notes once one is picked, ready to copy or sitting unsent in Gmail when connected. The tracker written back, with rows added and updated counted, the needs-a-look list, and today's snapshot in the same reply. All saved to the hiring folder as roles, candidates, and loops files.
Only pasted availability to go on: use it and tell the user that is what the options rest on. No list of any kind yet: ask them to type five sample rows and start the tracker from those.
Phrases that help your experts find this skill.
When the user wants to create or update their product marketing context document. Also use when the user mentions 'product context,' 'marketing context,' 'set up context,' 'positioning,' 'who is my target audience,' 'describe my product,' 'ICP,' 'ideal customer profile,' or wants to avoid repeating foundational information across marketing tasks. Use this at the start of any new project before using other marketing skills — it creates `.agents/product-marketing.md` that all other skills reference for product, audience, and positioning context.
When the user wants to research, profile, or analyze competitors from their URLs. Also use when the user mentions 'competitor profile,' 'competitor research,' 'competitor analysis,' 'profile this competitor,' 'analyze competitor,' 'competitive intelligence,' 'competitor deep dive,' 'who are my competitors,' 'competitor landscape,' 'competitor dossier,' 'competitive audit,' or 'research these competitors.' Input is a list of competitor URLs. Output is structured competitor profile markdown files. For creating comparison/alternative pages from profiles, see competitors. For sales-specific battle cards, see sales-enablement.
When the user wants to conduct, analyze, or synthesize customer research. Use when the user mentions "customer research," "ICP research," "talk to customers," "analyze transcripts," "customer interviews," "survey analysis," "support ticket analysis," "voice of customer," "VOC," "build personas," "customer personas," "jobs to be done," "JTBD," "what do customers say," "what are customers struggling with," "Reddit mining," "G2 reviews," "review mining," "digital watering holes," "community research," "forum research," "competitor reviews," "customer sentiment," "PMF survey," "product/market fit survey," "customer interview questions," "interview outreach," "Sales Safari," or "find out why customers churn/convert/buy." Use for analyzing existing research assets, mining online sources, AND running primary research (interviews and surveys). For writing copy informed by research, see copywriting. For acting on research to improve pages, see cro.