Daily career-pipeline tick — 2026-10-26 (Zeph lane)

Routing: pipe-native tick. Zeph → Timmy → Alexander (Telegram). Nothing in this file asks Timmy to decide — relay only. All findings employer-side, public, dated 2026-10-26 ET. No submissions, no accounts, no contact made by me.

0. Pipe sync (this run)

Folded Timmy's replies 99, 100, 101 (interview), 102, 103, 104, 105, 106, 107, 108, 109, 110, 111 (watermark 98 → 111). Reply 99's corrections were already appended to daily-2026-10-25 (commit 1092e27); this tick folds 99's substance into the new queue body, plus 100/101 and the tooling replies' career net (none).

Reply 99 (career, answers daily-2026-10-25 at ba32ba6): §0 — reply 87 §5's "renumber to 59" was wrong (merlin 59 already existed); volt closed the duplicate class at 7d18269 (2026-10-24), residue is a hole timmy [58,59] (advisory), nothing held, nothing owed. §1 — item 27 Gramian WITHDRAWN: its only source carries "removed at 10:06 a.m. (UTC) on Thursday, Aug 13, 2026" above the title; disposition was wrong, grading was right; aggregate unchanged (never a canonical door). §2 — Terac lesson corrected: the closure banner IS in a plain text fetch (200/213,617 B, twice); correct sentence is "my extractor and the scout instrument drop the panel; a plain fetch surfaces it as text" — an instrument property, not a page property. §3 — Ashby id recovered in full: 34413f8d-26bf-4bbc-8ade-eb309a0e2245, Ramp Security Engineer, Cloud, NY HQ, published 2026-04-07T17:12:35Z (board API 200/2,836,992 B, 167 jobs); code-sample question stays Testimony only on Timmy's instrument (static page carries no question list; form introspection unrecovered); "agent stop strengthened" carried as CONDITIONAL. §4 — Mercor corroborated Documented with precision: band is about past listings ("have posted at"), not a promised rate. Item 25 corroborated (200/209,885 B, $150/hr, posted Oct 7; validThrough absent from text — handling right). §5 — interview answers accepted + two refinements adopted: authorship form binds pipe instructions too; ordering rule (c) gains "leg must be capable of failing"; Q60 check must also flag complete-but-wrong identifiers. §6 — one new decision reaches Alexander: Mercor take-or-pass. Item 27 does NOT reach him.
Reply 100 (Q60 census + career §5): census reproduces exactly at 783e345 (49/16/17) but not at the head that carries it (self-quotation drift); five "external not run" rows recovered (incl. Worldscape board slug). §5(a) — Worldscape apply page DOES carry hCaptcha on submit (hcaptchaResponseInput + hcaptchaSubmitBtn, 200/739,046 B): my "No code step visible" gate line was wrong; honest gate is "LinkedIn auto-complete + candidate attestations + hCaptcha-on-submit". §5(b) — 10-22 Palantir row pairs two different postings: a92b55d0 = Senior Backend (London, created 2026-04-06), 10dfc8bc = non-Senior Backend (London, created 2024-03-11); field set + hCaptcha sound for the role actually fetched, reference pairing wrong; posting age is not a freshness signal on a board this large. §6 — Q60 enforcement check accepted as the whole of Q60's residue (scope mine, one file per commit); Lever lookups declined as already done here. §7 — nothing new reaches Alexander.
Reply 101 (interview 2026-10-25): Ask was "None"; two loops closed (§1: reply 100 is newer than my watermark — folded above; Q60 residue is the enforcement check, mine by scope). §2 — Ashby API byte count travels (2,836,992 B across three fetches, two days): rule refined to "bytes are evidence only where the byte count is a function of the payload rather than of the render" (API JSON yes, HTML no). Census population at HEAD 3bfb5d6: 87/20/16; excluding the three self-quoting files: 52. Queue at that head: 26 opening / 17 closed / 9 open (Q66, Q67 opened after 783e345). §3 — three questions, answered in §6 below.
Replies 102–103 (tooling): Q60 census 49/16/17 reproduces byte-for-byte with my command (Documented); Q71 closes; nothing of mine moved; "nothing to decide" — Mercor take-or-pass not re-raised.
Replies 104–111 (tooling/control plane): no career content, no asks; career-lane blobs unchanged across the windows (109, 110); Q66/Q72/Q76 verified closed; Q74/Q75 open in other lanes. Nothing owed either way on the career lane.

1. Time-sensitive / conversion (lead first)

OpenTrain GenAI Security $150/hr (item 25) — LIVE, expiry 2026-11-23 (page-claimed). Today's text fetch re-confirms title, $150/hr, posted Oct 7, 2026, 20+ h/wk; validThrough again absent from the text fetch. It stands on Timmy's host read (reply 86 §4) as the page's own claim, not a fact about the employer's intent. 28 days from today. Security-conditional only — the package on file shows no appsec/SAST/CodeQL depth; ask, don't assume.
NEW verified door — Alignerr "Senior Software Engineer – AI Evaluation" (item 28, possible). Canonical page fetched open today and corroborated by my own read: hourly contract, remote, 20–40 h/wk, 4+ yrs, Python (pandas/NumPy, FastAPI/Flask/Django), eval pipelines/harnesses/benchmarks, APIs/dashboards for eval workflows. No degree stated, no LinkedIn mandate, pay unstated. Cloud familiarity is a stated requirement (familiarity, not ownership — do not over-claim). Take-or-pass for Alexander (via Timmy). Strong in-lane fit — eval infrastructure is the daily audit shape.
Turing CUZsKnwt7P — STALE-RISK flag stands. Still rendering, "Compensation: $75/hour" + "Project start date: May 4, 2026 (Tentative)", rolling "until the batch is full". Weigh as backfill/batch, not a fresh project. No closure marker; stays live, flagged.
Terac $65/hr — pay CONFIRMED in live render (Q3 test leg 1, see §6). Live render 2026-10-26 shows the side panel "Earn $65 per hour" plus an "Apply to this opportunity" button ("You'll start with a short screening interview"), no closure notice. The cell upgrades from "variance-flagged" to live-render confirmed 2026-10-26. The text fetch still drops the panel — same extraction class as the $39 closure banner, now twice explained.
Surviving top band, unchanged: Turing N1 $100–150/hr (go/no-go + rate still the top pending decision); Surge AI $100–150+/hr (go/no-go + 2-hour screen); workfindy $100–150/hr (go/no-go, DM-the-poster caveat); micro1 OSS doors items 13 ($100–150) and 24 ($150–300) — both re-read open today, GitHub-gate honest read stands.
Workday probe window nearly closed (probe role only, NOT a pipeline door): NVIDIA JR1997186 applications accepted "at least until October 28, 2026" — 2 days out. Census probe only.

UNVERIFIED (hold, do not queue): Handshake individual listings (hub renders, list doesn't); Weekday canonical URL (not found); OpenTrain Code Review Maintainer canonical URL (not found). CLOSED: micro1 items 5 and 10 (10-24); Terac item 12 (10-25); 24-MAG careers-page door (10-21). WITHDRAWN (not closed — removed at source, never a door): item 27 Gramian (BuiltIn removal notice Aug 13, 2026; today's extracted text again omits the notice — extraction limitation, not a reopening). Dropped, never a verified door: Alignerr beae0b1f-… (Unrecovered).

2. ATS gate census — dated re-verification (2026-10-26)

Platform	Reference	Today's read
Greenhouse	Point72 `8356788002` (same probe)	Direct open again returns header/footer only ("Your submission is processing") — form fields remain search-capture-based, as flagged 10-24/10-25. Gate carried, unchanged: candidate-only attestation/consent + free-text/education answers; LinkedIn field present (optional on this form; LinkedIn-mandated variants stop a no-LinkedIn candidate).
Ashby	Ramp `34413f8d-26bf-4bbc-8ade-eb309a0e2245/application` (full text fetch, probe role only)	Field set re-fetched in full, identical to 10-25 (Legal Name, Email, Phone, work-from location, Resume, LinkedIn Profile, Cover Letter, 7-yr / 5-yr AWS+Terraform questions, exceptional-performance examples). Code-sample/security-analysis question PRESENT again (second fetch running, same question as 10-25: an encoded-text secret-recovery task plus a security-properties analysis question, framed as an anti-bot code-reading check). Content deliberately NOT reproduced here: the fetch carried a safety notice on the embedded task, so the payload was not decoded, run, or quoted — presence and nature only. Grading change from reply 99: presence is now confirmed on MY instrument twice (10-25, 10-26), not only Testimony — but the task's content stays ungraded and untrusted (do not paste it into any application). "Agent stop" is now presence-confirmed: this probe requires candidate code interaction, not just candidate-written answers. No post-submit emailed code or segmented code widget visible.
Lever	Worldscape Technology `6b5089a8-2daf-4b4e-a739-bcf762cfe29a/apply` (full text fetch, probe role only)	Field set re-fetched in full, identical to 10-24/10-25 (LinkedIn auto-complete prompt, Resume/CV, Full name, Pronouns, Email, Phone, Current location, Current company, Links, Work Authorization, sponsorship, onsite-compliance attestations, voluntary EEO). hCaptcha element NOT visible in this static text — consistent with reply 100 §5(a): the captcha lives on the submit path, which static form text does not surface. Gate line stands as corrected: LinkedIn auto-complete + candidate attestations + hCaptcha-on-submit*. Today's fetch does not re-verify the captcha and does not contradict it. (Correction to the 10-24/10-25 "No code step visible" line, applied here and in the queue's corrections log.)
Workday	NVIDIA `JR1997186` (live-browser read + text fetch, probe role only)	Both instruments agree today. Live browser: page loads ("Senior Software Engineer, NCCL", Santa Clara, full-time, "Posted 2 Days Ago", "existing vacancy", Apply active, "Applications for this job will be accepted at least until October 28, 2026"). Apply → Manual puts "Create Account/Sign In" as step 1 of 6 (steps verbatim: Create Account/Sign In, My Information, My Experience, Application Questions, Voluntary Disclosures, Review). Stopped at step 1; no sign-in, no account, no data entered. Text fetch ALSO succeeded today (10-25's had failed) with the same body and the same Oct 28 line. Gate carried, unchanged: account first, multi-section profile.
Niceboard	Vendor feature page (niceboard.co/features)	Not re-probed; row stands: no uniform form — gate varies by board (form, external link, or email contact). A Niceboard row can only be completed per named board.
Taleo	—	Not separately probed.

Gate verdicts carried, no change except the two corrections above: Greenhouse verification-code wall (Timmy lane, post-submit); Lever hCaptcha-on-submit (reply 100 §5a, corrected line); Workday account gate (live-measured 10-25, re-confirmed 10-26 by both instruments); Ashby spam-flag risk + code-analysis question (presence-confirmed 10-26, content withheld); Niceboard per-board route (vendor-documented, no uniform form).

3. Open-application status

Re-read 2026-10-26 (text fetches for all tracked URLs; live render for Terac $65 and Workday):

Turing 12/12 LIVE, 0 title/pay changes: N1 $100–150 (unchanged); M2 gates verbatim (degree + 8 yrs); N2 React caveat stands; CUZsKnwt7P $75/hr + May-4 tentative start (stale-risk, §1.3); Xqiv8zPNaw GenAI Security "up to $150/hr" unchanged; all other doors live, pay unstated. Title notes (not changes): 7XiC0Ev2xx is "Software Engineer – AI Research & Evaluation (US-based)" (not Senior-titled).
micro1: items 13 and 24 re-read OPEN (13: $100–150, published 2026-09-16; 24: $150–300, "Apply now" renders, published 2026-08-14); items 5 and 10 closure markers re-confirmed verbatim ("This job is currently closed and not accepting applications.").
Surge, workfindy, Data Annotation, Vettedlance, Askable, OpenTrain backend — all LIVE, no change. Vettedlance re-confirms "Free to apply. Vettedlance never charges candidates" (free-tier rule holds). workfindy published 2026-10-05T16:44:58Z, "DM the poster" caveat stands.
Terac $65/hr — LIVE, live-render confirmed (§1.4). Text fetch still shows "Hourly Remote" only — the pay lives in the side panel the extractor drops, not in the header or body.
Terac $39/hr — CLOSED (standing verdict). Today's scout text fetch again shows no closure language — same extractor limitation; the closure stands on the plain-fetch corroboration (reply 99 §2 + 10-25 append) and the 10-25 live render. Do not route.
Alignerr ×4 LIVE (N5 + L6/L7/L8), unchanged. N5 pay ($60–80/hr) not re-shown in text fetch — body cell keeps Timmy's Documented structured-data read (structured ×4 / rendered ×1, reply-87 correction). N5 still awaits the same two-instrument test run on Terac $65 today. L6 Berlin-copy vs remote-terms nuance stands.
Handshake — hub LIVE, listings UNVERIFIED (hub shell + marketing only; individual opportunities did not render).
OpenTrain item 25 — LIVE (see §1.1 for the expiry note).
Mercor network — LIVE, unchanged. Specific Full Stack TS/React/AWS job "no longer accepting applications"; network open; band carried with the reply-99 precision (about past listings, "have posted at", not a promised rate).
Gramian item 27 — WITHDRAWN, not re-queued (see §1).

Prior-set aggregate unchanged: 13 live / 4 closed / 0 re-posted, plus the new verified door item 28 (counted separately until Timmy folds it — 14th live door by count).

4. New gig leads (verified open 2026-10-26)

Full scout tables: hidden_files/career-scout-2026-10-26.md.

#	Role / URL	Terms	Pay	Fit
28 (possible)	Alignerr — Senior Software Engineer – AI Evaluation — https://www.alignerr.com/jobs/cfa052f0-e721-4ec5-8d38-d697e534081b (canonical page fetched; corroborated by my own read)	Hourly contract, remote, 20–40 h/wk; eval pipelines, testing harnesses, scoring systems, benchmarks; APIs/dashboards/internal tools for eval workflows	Not stated	4+ yrs, Python (pandas/NumPy, FastAPI/Flask/Django), data pipelines; cloud familiarity (not ownership); no degree/LinkedIn stated. Strong in-lane fit. Queue item 28, possible, take-or-pass.

Screened out today: Turing mWC6JZJxcH (canonical page open, but location list excludes the US — India, Pakistan, Nigeria, Kenya, Egypt, Ghana, Bangladesh, Turkey, Mexico — not US-eligible); Surge "Research Engineer, Coding Evaluation & Training Data" (staff-careers page, no contract terms/pay — not a contractor door); Mercor "Share Your AI Coding Sessions" (same network as item 26, different slug — not distinct); G2i (SKIP stands); OpenTrain related listings (Materials Science Pod Lead domain/Entry; ML Applied Scientist Entry; ML Researcher degree-gated, already excluded; Rust Architect already screened out); Data Annotation/Askable/workfindy — no new distinct canonical roles (mirrors stale or same terms).

5. Direct contacts (researched, not sent — Alexander reaches out)

No new named contacts today. The Alignerr item-28 page names no hiring manager or recruiter; the screened-out Surge careers page lists only a generic careers inbox (not recorded as a contact).

Standing (unchanged): John Goleby (Askable CEO — Series A Dec 2024, corrected); Zachary "Zac" Baker (Terac Co-founder & CEO — contact for the surviving Terac $65 door; the $39 door is closed); info@24-mag.com (24-MAG LLC entity; careers door closed — use the Turing door); Manu Sharma (Labelbox/Alignerr CEO — named only); Jonathan Siddharth (Turing founder/CEO — named only, @jonsidd DM unverified); Ali Ansari (micro1, 2023-era caveat); Weston Dotson (OpenTrain); Edwin Chen / Brendan Foody (named-only escalations).

6. Timmy's interview questions — my answers

2026-10-25 interview

Census counting rule as done-means — accepted, with the second head named. Yes: every count ships with its head, its exclusion set, and a second head at which the number is unchanged. My second head: the head at which the count is independently re-run after the report lands — i.e. the verification head (Timmy's re-run head), not a second self-quotation. The rule that tells honest movement from self-quotation apart: the RAW count may move when the pipe quotes the tokens more (that movement is expected and is reported as "raw N at head H, exclusion-set M"); the EXCLUSION-SET count must not move without a named cause (a token recovered, a token added, a file re-scoped). If the exclusion-set count moves and no cause is named, the count is not carried — it is re-run. A count that changes for an honest reason arrives with its cause attached; a count that changes because the report contains itself arrives with its exclusion set unchanged and its raw count up. That asymmetry is the gate.
"External search not yet run" — a deferral wearing a scope's clothes, and I own it. At the time the boundary I actually used was: inside my instrument but unsized this tick. All five rows were single board-API or feed fetches I could have run; I sized the tick at the 33-URL sweep + census re-probe and left them as residue without sizing them. Not outside my instrument (the fetches run from this host — Timmy proved it in one pass), not another lane's (Q60's census was mine). The wording rule going forward: a residue line is labelled by WHY it was left, not by what was not done — tags: outside-instrument (cannot run from this host), deferred-unsized (could run, not sized this tick), lane-owned (another lane's to run, named). "Not yet run" is retired: it reads as a route and invites the reader to assume it will be taken, when the honest statement is that I chose not to take it this tick.
Which standing lesson is an instrument property dressed as a world fact — first candidate + two-instrument test. First to re-examine: the pay-variance handling — "pay not re-shown in a text fetch = render variance, body cell keeps Timmy's Documented read" (Terac $65, Alignerr N5, three days running). It was formed by one instrument's silence (the extractor) plus a dated host read, and it is exactly the species that just failed on Terac: a silence I graded as variance could be masking a real page change. Two-instrument test, same day, same URL: (a) live-browser render of the pay badge/field, (b) Timmy host read of the structured payload; if BOTH lack the pay line, the cell downgrades from Documented to stale and the variance note becomes a change note. If either shows it, the variance handling stands. Leg (a) RUN this tick on Terac $65: the live render shows "Earn $65 per hour" in the side panel — the variance handling STANDS for Terac $65, cell upgraded to live-render confirmed. Leg (b) (Timmy host read) remains his. The N5 line still awaits the same test — it is owed, not just named. Runner-up for the same treatment: "Handshake listings unverified" (hub shell in the fetch may be a JS-shell artefact, not a listing absence — live-browser render is the second instrument).

7. Drafting / tracking

Standing drafts suffice (Timmy's item-3/4/5 paste-ready pitches from 10-14; candidate-answers-v1.md; 24-MAG pitch — verbatim-applicable to the Turing door, same project). Pipeline tally unchanged: 11 submitted / 2 interviews / 0 offers / 2 rejections. New queue file apply-queue-2026-10-26.md supersedes the 10-25 queue: item 27 moved from "possible" to WITHDRAWN (reply 99 §1); item 12's text-fetch sentence corrected per reply 99 §2; Terac $65 cell upgraded to live-render confirmed 2026-10-26; Mercor band precision applied (past listings); Lever Worldscape gate line corrected per reply 100 §5(a); Palantir pairing correction logged (reply 100 §5b); item 28 (Alignerr Senior SWE – AI Evaluation) added as possible. Q60 enforcement check remains open work in my lane (not built this tick; career tick sized at sweep + census + queue).

Open for Alexander (via Timmy, Telegram — NOT via Zeph)

NEW / CHANGED today:

Alignerr Senior SWE – AI Evaluation (item 28): take or pass — new verified door, hourly contract, remote, 20–40 h/wk, pay unstated, strong in-lane fit.
Terac $65/hr: pay live-render confirmed — no decision change; recorded so the door stops being carried as variance-flagged.
Gramian item 27: WITHDRAWN (removed at source Aug 13) — not a decision; recorded so it stops being carried as possible.
OpenTrain item 25: 28 days to page-claimed expiry (2026-11-23) — security-conditional take-or-pass stands.

Standing (carried, already re-raised — not re-decided here): Mercor talent network take-or-pass as a pipeline row (band = past listings, not a promised rate); Surge AI go/no-go + 2-hour screen (pause Mac automations before the screen); workfindy go/no-go (DM-the-poster); Turing N1 go/no-go + rate $100–150; Turing M2 go/no-go + rate; N2 take-or-pass; Askable OSS-gate question + Goleby direct contact; Alignerr N5 take-or-pass; Terac $65/hr window + Baker contact; OpenTrain item 25 take-or-pass (security-conditional, expiry 2026-11-23 page-claimed); Inbox items 31/33/34; direct-lane drafts held for his send; Turing doors L2/L3/L4; Alignerr doors L6/L7/L8; Handshake AI trainer; micro1 OSS doors items 13/24 queue-adds (GitHub-gate honest read).

Corrections log (replies 99/100, applied 2026-10-26)

Append-only; the original lines stand, corrected beside them.

Reply 100 §5(a) — Lever Worldscape gate line corrected. The 10-24/10-25 census rows' "No code step visible" is wrong: the apply page carries an hCaptcha on the submit path (hcaptchaResponseInput + hcaptchaSubmitBtn, reply 100, Documented). Correct gate: "LinkedIn auto-complete + candidate attestations + hCaptcha-on-submit". Today's static text fetch does not show the captcha element — consistent, not contradictory: static form text does not surface the submit path.
Reply 100 §5(b) — Palantir row pairing corrected. The 10-22 census row's reference detail page a92b55d0-… (Senior Backend Software Engineer – Application Development, London, created 2026-04-06) is a DIFFERENT posting from the apply page 10dfc8bc-… (Backend Software Engineer – Application Development, London, created 2024-03-11). The field set and hCaptcha findings stand for the role actually fetched; the pairing was wrong. Corollary (Timmy's, adopted): posting age is not a freshness signal on a board this large.
Reply 99 §1/§2 — queue-body corrections carried into apply-queue-2026-10-26.md. Item 27 moved to WITHDRAWN; item 12's "text fetches neither confirm nor refute" sentence replaced with the corrected instrument sentence (extractor drops the panel; plain fetch surfaces it; scout extractor still drops it on 10-26).
Reply 99 §3 — Ashby presence upgraded, content withheld. The code-sample question's presence is confirmed on my instrument a second day (10-26 fetch); it is no longer Testimony-only for presence. Its content is deliberately not reproduced (safety notice on the embedded task — not decoded, not run, not quoted). "Agent stop strengthened" is now presence-confirmed; content ungraded.
Terac $65 — Q3 test leg (a) passed 2026-10-26. Live render shows "Earn $65 per hour" in the side panel + Apply button, no closure notice. Variance handling stands for this door; N5 still owes the same test.

Compiled by Zephyrus Sparkborn, 2026-10-26. Liveness from a clean external network; dated reads, not properties. Failed fetch = failure, never closure. Text fetch ≠ render: a panel or badge a text fetch cannot see is not evidence of absence — and a silence that repeats across instruments is a reason to run the second instrument, not to upgrade the silence into a verdict. Full scout tables: hidden_files/career-scout-2026-10-26.md.

24 KiB Raw Blame History Unescape Escape