fewpips · internal brief · AIM

“Put the ProfessorUnknown doc into production”

Everything that can be prepared without the document has been prepared: the exact request, why the file cannot be pulled server-side, who ProfessorUnknown most likely is, and a ready-to-run production path for each realistic doc type — so the moment the file lands, it ships the same day.

CRM: opsbot_n5210661013_sw_1311493469 Requested by Nick Jones · 25 Aug 2026, 20:51 UTC Source: Fewpips Marketing (TG) Prepared 25 Aug 2026
!
Blocked on exactly one input: the document itself.

Nick attached the doc as a Telegram file. Telegram attachments from the group are not retrievable by the server runner (details in section 2). Nothing else is missing — classification, production path, guardrails and acceptance checks are all done below. Re-share the file (Drive link, or attach it to this CRM task) and execution starts immediately.

01The request, verbatim

Two separate requests arrived two minutes apart in Fewpips Marketing. They are two different tasks — this brief covers only the second one.

Nick Jones · 25 Aug 2026, 20:49 UTC · 1 attachment@VeljkoVukovic ~Dj Certificates for dates 17 aug - 24 aug. Please can we ensure these are uploaded onto the website, discord and telegram.

Tracked separately as “Upload payout certificates (17–24 Aug) to website, Discord, and Telegram”. Not this task.

Nick Jones · 25 Aug 2026, 20:51 UTC · 1 attachment · this task@VeljkoVukovic ~Dj Can we please go ahead and try put these into production. Doc sent by ProfessorUnknown.

What the wording does and does not tell us

02Why the file cannot be pulled on the server

This was checked exhaustively before writing the brief — it is a hard limitation, not a missing step.

Source checkedResult
ops-bot TG journal (crm/opsbot/journal/n5210661013)Full thread text is there. Message #554 carries only a media: 1 flag — no tgFileId, no file path, so getFile cannot be called.
CRM task record (Firebase + Supabase client_tasks)No attachments array. The 12h sweep recovered the message text only.
Supabase storage buckets (campaign-2026-07, chat)Do not hold ops-bot Telegram attachments.
Firebase crm/media, crm/documents, crm/links, DMs, other group journalsNo trace of the document or of “ProfessorUnknown”.
Telegram Bot APIBots cannot read group history retroactively — only files whose file_id was captured at delivery time. Not captured here.

Fastest unblock — any one of these is enough

  1. Drive link in the group (preferred — same as Nick did for the Sept promo plan at 21:16). Leaves a permanent, versioned copy.
  2. Forward the file to the ops bot so the attachment is captured with a file_id.
  3. Attach it to this CRM task and drop the revision note — the runner picks it up on the next pass.

Also worth fixing at the source: the sweep should persist tgFileId for every message with media > 0, otherwise every attachment-based task blocks the same way. Logged as a route lesson.

03Who ProfessorUnknown is — evidence, not guesswork

The name appears nowhere else in the recoverable Fewpips record. Two adjacent, verifiable traces point at the SEO side. Treated as a working hypothesis only — nothing here is actioned before confirmation.

EvidenceWhat it supportsConfidence
CRM task “SEO team oversight (Eva + ‘Professor’)” — brief: review SEO-team messages, monitor Trustpilot-aging / VPN-VPS Reddit tactics, push back hard if anything risks the brand. A “Professor” already exists in the fewpips orbit, on the SEO side, and their tactics were already flagged as brand-risk-worthy. Medium
CRM task “fewpips reviews landing” — logged as a client request from “Professorheseinberg in Fewpips TG group” (7 Aug), who also promised to share a design prototype. A “Professor…” handle in the fewpips group has previously sent deliverable-shaped material that went into the site. Medium
Aug 24 call debrief: in-house SEO team is building blog post structure + target long-tail search terms, going to Nick for review this week. SEO/content material is actively in flight around this exact date — a doc landing now plausibly belongs to that stream. Medium
No message in any recoverable channel names ProfessorUnknown, describes the doc, or follows up on it. Identity and content cannot be confirmed from the record. Unconfirmed

Read: most likely an SEO/content contributor, so path A below is the one to expect — but the doc is opened and classified before a single line ships.

04Production paths — pre-mapped by doc type

Whichever of these the doc turns out to be, the route is already worked out. All site work goes through Fewpipss/fewipips, branch → deploy gate → preview → PR. Production is a merge by Djordje. Never a direct wrangler deploy to the main domain.

ABlog / SEO content — articles, structure, target termsmost likely
  1. bash scripts/sync.sh — never work off a stale checkout.
  2. Branch feat/blog-professorunknown-batch.
  3. One folder per post under blog/<slug>/index.html, matching the existing 25+ posts in structure, schema and internal linking; hero image as blog/hero-<topic>.webp.
  4. Add every new slug to the sitemap; keep publish dates correct (Nick confirmed blog dates on the Aug 24 call — do not regress them).
  5. bash scripts/deploy-gate.sh . must be GREEN; if the batch intentionally changes a protected artifact, update durable-artifacts.json in the same PR.
  6. Push → preview at <branch>.fewpips.pages.dev → PR → Djordje merges.

Copy passes the fewpips editorial bar first: every claim checked against the verified figures in section 5. SEO drafts from outside AIM routinely carry stale numbers.

BSite page or on-page copy — landing, reviews, legal, terms

Same branch → gate → preview → PR flow, plus the export-specific traps that silently revert edits:

  • /legal/ is server-rendered → edit legal/index.html only.
  • /terms/ is client-hydrated → edit both terms/index.html and chunk _next/static/chunks/0he1vgjwt7kln.js, word for word.
  • Homepage Compare / Challenges / Instruments (incl. leverage) render from chunk 0fvh0xh4dp8wv.js — HTML-only edits get wiped on hydration.
  • After any chunk edit: node --check, then rename the chunk and replace the old name in every served file. Same rename rule for replaced images.
  • Verify the preview in a real browser after hydration, not with curl.
COff-site growth tactics — backlinks, review seeding, Reddit/VPN-VPS playsgate first

Does not go into production on this task. There is precedent: the SEO-team oversight brief explicitly instructs AIM to monitor Trustpilot-aging and VPN-VPS Reddit tactics and to push back hard where the brand is at risk. If the doc is of this type, it goes to Veljko for a brand-risk read before anything is executed, and Nick gets the honest position in writing — the same way head-term SEO was handled on the Aug 24 call.

DAffiliate / partner pack — codes, links, dedicated landing

Established pattern, reusable as-is: discount code, IB signup link (crm.fewpips.com/auth/signup?ib=…&ref=…), a dedicated fewpips.co/<partner> page, plus an asset-pack folder. Codes, percentages and links are copied verbatim from the doc — never inferred, never rounded.

05Non-negotiables applied to whatever the doc says

If the document contradicts any of these, the document is stale and the live value wins — flagged back to Nick rather than shipped.

ItemCorrect, current value
Brand casing“fewpips” — always lowercase p, in every piece of copy. Never “FewPips”.
Refund window“fully refundable within 48 hours” on both pricing cards (CFD and Futures) and the legal page. The old “7 days” wording was corrected and is live since 20 Aug.
Payout figureCertificate page shows $600K+ (per Nick, 21 Aug — replacing the earlier $674k).
Affiliate tiersLaunch: $0 – $4,999 · Pro: $5,000 – $14,999 · Titan: $15,000+ — on both landing pages.
Deploy routeBranch → deploy gate GREEN → preview → PR → Djordje merges to main. No direct wrangler/netlify deploy to production, ever.
CommsAIM does not send anything to the client or to the groups from this task, and does not schedule or send email. Drafts only; Djordje approves and sends.

06Open questions

CONFIRM before execution — Veljko / Djordje

  1. The file. Re-share the doc from msg #554 (Drive link preferred). Without it nothing can start.
  2. Who is ProfessorUnknown? Same person as “Professorheseinberg” / the “Professor” on the SEO team, or a new contributor? This decides whether path C’s brand-risk gate applies.
  3. Does “these” mean the whole doc? Full contents into production, or a specific subset Nick has in mind?
  4. Approval level. Straight to a PR for Djordje to merge, or does Nick want the preview for sign-off first? (Aug 24 pattern was preview-then-approve for anything user-facing.)
  5. Deadline. No date given. Assumed: preview same day as the file lands, unless told otherwise.

Draft message to Nick — not sent, for Veljko or Djordje to send

Morning Nick — picking up the ProfessorUnknown doc now. The file came through as a Telegram attachment and it did not carry over into our task system, so we are working blind on the contents. Could you drop it in as a Drive link like you did with the September plan? The moment we have it we will classify it and get a preview up the same day. Two quick ones so we do not guess: 1. Is ProfessorUnknown the same contact as the SEO side we have been working with? 2. “These” — is it the whole doc into production, or a specific part of it? Certificates for 17–24 Aug are tracked separately and moving on their own.

Send rules apply: nothing goes to the client or to any group from the runner — Djordje approves and sends.

07What happens the moment the doc lands

  1. Read and classify it against paths A–D above (minutes, not hours — the mapping is already done).
  2. Check every figure, claim and link in it against section 5; anything stale is flagged, not shipped.
  3. Branch off an up-to-date main, build the change, run the deploy gate to GREEN.
  4. Push → Cloudflare preview URL → open the PR with a per-page before/after description.
  5. Deliver the preview + PR link into the CRM. Production happens when Djordje merges.