# Portfolio process record October 5, 2026. Homepage shipped in PR 88. Original wording is preserved below. # Curated decisions ## Start with the actual portfolio Sep 30 Request: Style the existing case study from the portfolio and résumé. Response: Run the real site; use the supplied Figma Slides file as the design target. Outcome: Source material anchors the work. ## Change hierarchy after polish Sep 30 Request: Headings and text still feel wrong; do more layout and art direction. Response: Six recorded polish and art-direction passes across 44 active slides. Outcome: The composition and reading rhythm change, not only the surface style. ## Make the Staff story legible Sep 30 Request: Build a general sizzle portfolio for a hiring manager; represent Staff level. Response: A 13-slide portfolio prioritizes recent work, scope, ownership, and outcomes. Outcome: A separate 18-slide working edition adds navigation and marked evidence gaps. ## Make review possible away from home Oct 02, 19:05 Request: Create five header directions from main; then ask for a hosted interactive preview. Response: Signature, Systems in motion, Field notes, Work first, Design × engineering. Outcome: Hosted review replaces a local-only handoff. ## Carry the identity through exploration Oct 02, 19:41 Request: Likes 1, 2, 5. Keep recessed nav, the name at the top, existing content, and lightweight ASCII controls. Response: Five more: Studio, Masthead, Afterhours, Gallery, Margin. Outcome: Ten directions are printed into Paper, with desktop and mobile studies. ## Explore below the hero Oct 02, 21:39 Request: Ten options, subagents and a jury; consider interaction design. Response: Ten about-section layouts and five interaction studies. Outcome: The review expands from static composition to behavior. ## Turn comments into a new brief Oct 02, 22:57 Request: Praise clean biography and credentials; criticize small headings and color; request a letter interaction. Response: Three actionable comment edits are reported, then ten new directions after the clarification. Outcome: Facing pages, The two-column record, and Ink & daylight are the jury shortlist, not final user selections. ## Refresh the design source from main Oct 03, 21:46 Request: Pull the latest GitHub code and update options on a new Paper page. Response: The delivery records main at 8fd9f9b and ten updated studies. Outcome: Earlier review pages are preserved. ## A valid link did not prove ownership Oct 03, 22:03 Request: The file is not visible in the expected account. Response: The assistant acknowledges checking the link without checking ownership. Outcome: Original ownership resolution is not established in the preserved record; later work uses the user-supplied favorite file. ## Fit a header to the preferred body Oct 04, 10:24 Request: The supplied From the Notes composition is the favorite; show how the header fits. Response: Compact, Name-led, and Orbit stage studies, including mobile and ASCII-to-3D states. Outcome: The header/body relationship becomes the object of review. ## Use real media in the coded preview Oct 04, 15:58 Request: Create a preview branch; include real photos and vectors. Response: A coded name-led homepage reuses the live media collections. Outcome: Still-image approval becomes an interactive review. ## Make the navigation a site component Oct 04, 16:23 Request: The new navigation does not work across other pages. Response: Move the recessed navigation into the shared layout. Add four 28px visual treatments. Outcome: User selects option 3, Quiet index; SVG and rcweston remain in the identity. ## Keep context and social links Oct 04, 18:16 Request: Retain where I am at and socials; make the lockup feel designed; imply current status with a dot. Response: Restore role/socials, compose the signature, then replace the label with a status pulse. Outcome: The content is retained while hierarchy remains under review. ## Test the name and character together Oct 04, 18:42 Request: The dual stack does not work; explore three layouts plus three including the character. Response: Six full header lockups. Outcome: The settings button becomes the next unresolved relationship. ## Reject variations of the same lockup Oct 04, 18:56 Request: The settings button floats. Ten control options still feel like the same composition. Response: Ten complete header structures follow, carrying aspects liked in studies 2 and 4. Outcome: Exploration changes the underlying layout, not only button placement. ## Make the controls local and unbranded Oct 04, 19:54 Request: Do not label the object Orbit; use a button and the inline interaction rather than a modal. Response: Controls unfold within the frame, with responsive packing and keyboard dismissal. Outcome: Internal code names remain; visible controls describe actions. ## Explore three promising families Oct 04, 20:38 Request: Iterate ten times each on option 2, W2, and W3. Response: Twelve round-three studies lead to thirty round-four variations. Outcome: Assistant shortlist: 02-06, W2-07, W3-01. This is not the eventual selection. ## Select the photographic contact strip Oct 04, 22:20 Request: Move forward with W3-09; preserve all options in a repo folder under admin. Response: The homepage gets a dedicated ContactStripHeader; the comparison implementations stay available. Outcome: W3-09 is the selected preview direction. ## Freeze the larger composition Oct 04, 22:33 Request: Ask for two independent media players, then pause broader home layout changes. Response: Subsequent edits are localized to content, controls, alignment, and media settings. Outcome: The final preview has one Shots/Morphs reel plus three header photos. Two independent players are not delivered. ## Make the photographs maintainable Oct 04, 22:43 Request: Round corners like the character frame; allow Shots selection or a custom upload from admin. Response: Three configurable slots, authenticated save/upload handlers, image validation and per-slot fallbacks. Outcome: Code is implemented. Signed-in upload/save browser verification remains outstanding in the delivery record. ## Reduce the media chrome Oct 04, 22:44 Request: Call modes Shots and Morphs, default to Morphs, remove the counter and caption. Response: The mode-switchable reel is retained; visible counters and redundant overlay text are removed. Outcome: Accessible slide descriptions and playback controls remain in code. ## Select Compact stack Oct 04, 23:06 Request: Four role options; NYC can help. Choose option 1 and align the height. Response: Compact stack replaces the prior role line; logo, role and location share a vertical structure. Outcome: User choice differs from the assistant recommendation of Role first. ## Remove secondary interactions and duplicate content Oct 04, 23:14 Request: Remove the letterform/sign-off, optics caption, and lower social line. Move How I work into smaller bullets. Response: The illustration retains a recessed play/pause control; the work bullets follow the biography. Outcome: The page has fewer competing blocks. ## Correct the labels and experience dates Oct 04, 23:16 Request: Use Education; JLL 2016–2019; Traversal 2025–2026. Rename the chart Exploration. Response: Apply the later explicit correction over the earlier instruction to match the reference exactly. Outcome: The decision record preserves both the initial direction and its correction. ## Use one grid and a quieter principle marker Oct 04, 23:36 Request: Align headings to the paragraph column, add vertical space, improve company/education hover states, use Ethos and hexagons. Response: One shared grid spans biography, credentials, activity and Ethos; the principles use two columns at desktop. Outcome: This is the final selected preview checkpoint at a147b2d. ## Release the approved homepage Oct 05, 00:59 Request: Ship it; the case-study deployment did not include the homepage. Response: Preserve the published article while syncing and merging homepage PR 88. Outcome: Production deployment Ready; the custom domain visibly shows W3-09, Morphs, corrected dates, and the shared lower-page grid. # Archive checkpoints ## Start with the real portfolio 2026-09-30T21:12:28.604Z Prompt: Match the deck to the portfolio and use the résumé as evidence. Decision: Run the actual Portfolio 2026 project locally. Resolve the initial Paper/Figma ambiguity to the supplied Figma Slides case study. Outcome: The existing site and case study became the visual and factual starting points. This was an adaptation of Charlie’s work, not a fabricated portfolio. Learning: A source-faithful brief is more useful than a generic premium style. ## Polish exposed a hierarchy problem 2026-09-30T21:33:23.449Z Prompt: Premium and world class—then: “I don’t like the headings and text styles.” Decision: After the first polish passes, change composition and hierarchy: quieter sentence-case headings, larger product views, asymmetric layouts and dark chapter breaks. Outcome: 44 active slides received successive typography and art-direction passes. Before, polish and art-direction renders remain available as separate evidence. Learning: More styling did not resolve the structural concern. The response had to change where attention went. ## Make the work screen at Staff level 2026-09-30T22:25:46.776Z Prompt: Create a general sizzle deck that answers a hiring manager’s questions. Decision: Prioritize recent, consequential work and clarify role, ownership, complexity and outcomes using the published résumé. Keep all publication in the personal Portfolio workspace. Outcome: A 13-slide portfolio brings Arca, Traversal, Instabase and Atlassian into one Staff-level narrative. The personal-workspace rule became a persistent chat preference. Learning: Seniority is demonstrated through decisions and scope; it cannot be supplied by visual polish or invented metrics. ## Show the evidence still needed 2026-09-30T22:55:05.992Z Prompt: Make a more interactive version and indicate where more information, images or video would help. Decision: Use a clickable menu, before/after storytelling and visibly marked evidence gaps. Distinguish proposed media from material already available. Outcome: An 18-slide working edition preserves the polished story while exposing what must still be supplied. Learning: Visible gaps make a working portfolio actionable without pretending the evidence is complete. ## Explore the opening screen 2026-10-02T23:05:25.629Z Prompt: Create 4–5 homepage header directions on a branch from main; make them viewable away from the computer. Decision: Keep the exploration additive and build distinct compositions: Signature, Systems in motion, Field notes, Work first and Design × engineering. Outcome: Five directions, a hosted personal preview and 25 recorded viewport checks. This thread ran in parallel after Orbit branched. Learning: Review access is part of the design workflow: a local-only page cannot help someone away from their computer. ## Keep the name and recessed navigation 2026-10-02T23:41:01.142Z Prompt: Develop favorites 1, 2 and 5; retain the pushed-in navigation, the name at the top, ASCII controls and existing content. Decision: Add Studio, Masthead, Afterhours, Gallery and Margin. Treat the name and existing content as constraints rather than disposable layout elements. Outcome: Ten directions were printed into Paper with desktop and mobile reviews. Learning: A visual experiment can vary composition without erasing what makes a personal portfolio identifiable. ## Move from layouts into interaction 2026-10-03T01:39:59.818Z Prompt: Riff on the section in ten ways, use a jury, and make it feel more designed through interaction. Decision: Build ten about-section layouts and five interaction studies. Use separate reviewers to challenge composition and behavior. Outcome: Editable Paper review material and local interactive studies were retained alongside the static layouts. Learning: The interaction needs to support the content’s reading order, not just decorate it. ## Make the next ten answer the comments 2026-10-03T02:57:50.550Z Prompt: Read the comments—then generate ten more from that feedback. Decision: Resolve the actionable comments while retaining favorites as open signals. Expand into materially different typographic compositions. Outcome: Facing pages, The two-column record and Ink & daylight emerged as strong options. The recorded 50 viewport checks found no overflow or missing assets. Learning: Comments can be a new design brief, not merely a punch list. ## Ten passes, one silhouette 2026-10-02T23:06:14.420Z Prompt: Model a buddy in Blender through ten cycles and establish how it could behave on the website. Decision: Refine sphere → pebble → soft square; add gesture, grounding and manufactured detail; then remove excess expression and resolve articulated hands. Outcome: Ten native Blender versions, still renders and a web interaction study. The tenth was selected within that initial round, before Charlie’s later rejection of the range. Learning: Iteration depth and conceptual breadth are different. A highly resolved form can still answer the wrong question. ## The rejection that changed the brief 2026-10-02T23:43:41.330Z Prompt: “It also looks like it’s the same blob just iterations on him.” Decision: Fix review visibility and stop only refining the blob. Explore distinct families: Cursor, Folio, Loop, Orbit, Pixel and Spark. Outcome: A genuine silhouette comparison made selection possible remotely. Learning: The critical feedback was about sameness, not surface finish. The next round needed different underlying ideas. ## Let the logo determine the character 2026-10-02T23:51:18.260Z Prompt: Use the hexagon lockup. Keep the orb. Decision: Develop a logo-derived hexagonal shell and offset amber accent around an orbital core. Outcome: The early Hex Orbit model established the brand relationship that survived later character and motion revisions. Learning: A recognizable source of form made the character more specific to Charlie. ## Explore a family of moving objects 2026-10-02T23:59:18.823Z Prompt: Ten variations; Guilty Spark as inspiration; make them move. Decision: Build Keeper, Lockstep, Bloom, Gyro, Crescent, Facet, Relay, Iris, Lantern and Keystone with different shell and optical arrangements. Outcome: Ten native scenes, exported web models and individual motion clips. The family provided options for the deeper character pass. Learning: Breadth improved, but synchronized loops and target-like eyes still made the objects feel more like products than characters. ## Build acting, not another idle loop 2026-10-03T00:21:37.503Z Prompt: Spend more time. Aim for Pixar quality. Have a judge and jury. Decision: Use separate character and motion jurors plus a judge. Rebuild optics, lids and pivots; author notice, anticipation, curiosity, acknowledgment and recovery. Outcome: Eight craft rounds addressed anatomical-looking optics, weak gaze, lid penetration, oversized reflections and a rear chassis intersection. The final native film is 180 frames at 24 fps. Learning: The jury passed a refined portfolio prototype, explicitly not Pixar equivalence. Character 7.5/10 and motion 7/10 are the reviewers’ recorded judgments, not objective quality metrics. ## Judge it at the size people will see 2026-10-03T01:26:02.840Z Prompt: Render it on the website and record the interaction. Decision: Integrate the model into the real homepage, adjust the browser presentation and record actual browser frames. Outcome: Homepage screenshots, an interaction recording, raw frames and source were retained. Replaced recording versions were recovered from Git. Learning: A studio render is not evidence of how the character reads inside the portfolio. ## Give the transformation some resistance 2026-10-03T02:10:35.044Z Prompt: How does the ASCII blob become this? Make it “iron man itself together”—less clean. Decision: Stage the transition as independent shell assembly, with uneven timing, exposed mechanics and visible latching instead of a clean substitution. Outcome: The transition gained a storyboard and a browser implementation. Character parts assemble around the developing core. Learning: The premium feel came from a believable sequence of events, not a smooth crossfade. ## Repair the jump between the first two frames 2026-10-03T02:51:44.042Z Prompt: Add frames between 1 and 2; the latter half looks okay. Decision: Focus the change on the opening: add intermediate views of the material taking shape without restarting the already accepted second half. Outcome: Storyboard 1a, 1b, 1c and 1d make the early assembly progression inspectable. Learning: Precise feedback can narrow effort. Preserve what is working while making the weak transition legible. ## ASCII becomes a state worth staying in 2026-10-03T03:34:58.563Z Prompt: Follow the cursor. Activate deliberately. Let symbols evolve, congeal, spark and fight. Default to ASCII, including loading. Decision: Expand the glyph sequence before any material appears. Add bounded gaze and interaction, and preserve an alive ASCII resting state. Outcome: The documented sequence gives ASCII 7.2 seconds before a 4.25-second shell assembly. Activation is deliberate; the experience no longer races directly into 3D. Learning: The beginning became an experience in its own right, which made the transition feel earned. ## Make the behavior controllable 2026-10-03T04:19:57.960Z Prompt: Audit accessibility, then create a PR. Decision: Provide keyboard and non-drag alternatives, reduced-motion behavior, explicit pause and motion controls, and screen-reader names. Verify the interaction before preparing PR #82. Outcome: The review covers the Orbit experience and exercised flows; it is not whole-site WCAG certification. The branch and evidence remain available for review. Learning: Accessibility changed the interaction model as well as the labels. ## Turn text links into tactile controls 2026-10-03T13:24:46.251Z Prompt: Small icon buttons with the pushed-in aesthetic; screenshots and a Paper legend to review. Decision: Carry the recessed navigation language into compact outline icons while retaining usable touch targets, names, hover labels and keyboard focus. Outcome: Desktop/mobile state screenshots and a default-theme Paper review with a button/interaction legend made the controls concrete enough to critique. Learning: A smaller visual footprint does not require a smaller interaction target. ## Separate appearance, motion and playback 2026-10-03T20:04:44.712Z Prompt: Keep buttons from jumping. Motion off must freeze Orbit rather than revert to ASCII. Confirm mobile and add optional sound. Decision: Reserve fixed control positions, separate per-model motion from appearance, and distinguish a one-shot playback pause from all motion. Add an explicit mute control with sound off by default. Outcome: The assembled model remains visible when frozen. Mobile layouts and playback states were exercised; sound is gesture-enabled and suspended when appropriate. Learning: A control should change only the state its label promises to change. ## Reduce the resting visual noise 2026-10-03T20:29:50.533Z Prompt: Hide the floating buttons behind progressive disclosure, soften contrast and improve the design using component references. Decision: Develop a recessed expandable rail, restrained enabled icon contrast and explicit focus/hover labels. Use the references as interaction input rather than a substitute for the portfolio’s visual system. Outcome: The rail presents one resting entry point and reveals the controls together. Closed controls are removed from interaction. Learning: Progressive disclosure made it possible to keep the controls legible when needed without dominating the hero. ## Anchor everything to the center 2026-10-03T21:24:47.763Z Prompt: Put the X in the center. All icons should be visible when open, with unavailable actions shown as disabled strokes. Confirm AA rules. Decision: Fix the center control in place. Reveal appearance and sound to its left, motion and playback to its right. Keep the unavailable playback icon visible and explain its state. Outcome: The saved review measured enabled icon contrast at a minimum 3.24:1 and tooltip text at a minimum 9.93:1 across checked themes. Targets remain 44 px. Disabled controls do not imply an available action. Learning: Stable geometry makes changing states easier to understand. These measured checks are component evidence, not a blanket WCAG conformance claim. ## Ship the reviewed branch 2026-10-03T21:49:56.354Z Prompt: Deploy. Decision: Build a clean export of commit 0ce125a in the existing personal Vercel project, validate the staged deployment, then promote it to rcweston.com. Outcome: Production build and TypeScript passed. The deployed renderer and model matched the source bytes; live activation, playback pause/resume and assembled motion freeze were exercised. PR #82 remains unmerged. Learning: A release ends this documented round, not the opportunity for further refinement. # Homepage requests ## H19 · 2026-10-02T23:05:25.629Z create a new branch on main and i want you to give 4-5 directions for the home page header ## H292 · 2026-10-02T23:30:31.817Z I’m not at home ## H307 · 2026-10-02T23:31:10.641Z Question: Would you prefer a Figma review in your personal Portfolio workspace, or a hosted preview where you can try the interactions? Answer: Hosted interactive preview ## H406 · 2026-10-02T23:41:01.142Z 1-2-5 are the most interesting I want to keep the ui feel of the nav being pushed in so that should carry over as this design scales. I also want to keep all the content in home ( below the hero image ) as is today. The layout can adjust but content should carry over and feel free to add. On 5 I really liked the idea of using my rotating ascii thing as the object and giving some light weight controls to customize Largely explore this with 5 more directions. Make this not feel vibe coded ## H438 · 2026-10-02T23:42:29.207Z Keep my name at the top. It’s a portfolio landing page for me ## H851 · 2026-10-03T00:57:45.730Z Print every exploration in a new paper file ## H1430 · 2026-10-03T01:39:59.818Z Riff on this section and give ten options of varying design ## H1461 · 2026-10-03T01:41:33.449Z Spend subagents and jury on this. Paste to paper ## H1874 · 2026-10-03T02:09:55.277Z Think about how this can feel more designed ## H1876 · 2026-10-03T02:09:55.287Z Interaction design ## H2241 · 2026-10-03T02:57:50.550Z I left comments ## H2529 · 2026-10-03T03:25:54.179Z Sorry. I want you to take my feedback and generate 10 more ## H2905 · 2026-10-04T01:46:14.790Z Pull down the latest from GitHub and update the options in paper on a new page ## H3177 · 2026-10-04T02:03:39.751Z Where in paper is this ## H3189 · 2026-10-04T02:05:22.190Z It’s not there ## H3371 · 2026-10-04T02:24:11.832Z youre making this in a new paper account - not mine ## H3380 · 2026-10-04T02:24:20.376Z [https://app.paper.design/file/01M3ZWZPXAEM36M054ES0C06KP/p-2-0](https://app.paper.design/file/01M3ZWZPXAEM36M054ES0C06KP/p-2-0) this is the link ## H3521 · 2026-10-04T14:24:53.938Z this is my favorite compliantion [https://app.paper.design/file/01M42BH9BB1W9TMRKD11PD9ZTD/p-2-0/38T-1](https://app.paper.design/file/01M42BH9BB1W9TMRKD11PD9ZTD/p-2-0/38T-1) but i want to see how the header fits on thsi ## H4150 · 2026-10-04T19:58:02.036Z ok make a preview branc for me to review ## H4169 · 2026-10-04T19:58:52.960Z actually - i would like to have teh show real ( photos and vectors ) included - can you confirm thsoe fit ## H4487 · 2026-10-04T20:23:11.083Z this is better but we still need to figure out the top heading section and how that new nav works across other pages ( it doesnt right now ) ## H4953 · 2026-10-04T20:56:27.596Z i want the nav inputs more minimal ( 28px? ) and for the ui to feel more pushed in - i also want my svg to still be in the top left hand corner and for it to read rcweston next to it ( at least before scroll ) ## H4976 · 2026-10-04T20:59:59.019Z open to ideas here and woud like to see concepts ## H5268 · 2026-10-04T21:11:09.555Z option 3 ## H5407 · 2026-10-04T22:16:43.682Z i want to keep the 'where i am at' and social icons in the hader ## H5529 · 2026-10-04T22:26:41.638Z make this lock up feel 'designed ## H5539 · 2026-10-04T22:26:47.838Z [Image-only feedback; described in the linked decision record.] ## H5684 · 2026-10-04T22:36:03.105Z Do something different for 'currently' like a pin or flashing dot .. i dont want to say 'here' but infer it ## H5830 · 2026-10-04T22:42:51.453Z I feel like the dual stack doesnt work - can you mock 3 options of the layout ... i would ideally like to get the orbit icon in here as well - do three more explorations to cover that ## H6032 · 2026-10-04T22:56:12.860Z i like thes all but a big issue is that the settings button doesnt fit in the layout - it feels like it floats, explore 10 options into how this can be achieved better ## H6261 · 2026-10-04T23:45:29.775Z i hate all of this - mainly bc its the same lock up ## H6272 · 2026-10-04T23:45:48.945Z 2 is interesting in the designed nature of it ## H6306 · 2026-10-04T23:54:54.243Z 4 was cool ## H6308 · 2026-10-04T23:54:54.248Z also dont call it orbit - i dont want it branded for the end user - just a button for the controls to appear - i also dont like the modal - i like the nice ixd we came up with ## H6823 · 2026-10-05T01:46:48.375Z iterate 10 times on the option in 2, w2, and w3 ( 10 each ) ## H7230 · 2026-10-05T02:20:17.661Z keep all of these options in a repo folder on admin but lets move forward with w3-o9 ## H7432 · 2026-10-05T02:33:43.348Z ok - so i want this to be just shots and vectors ( include toggles to play and advance ) on each ## H7464 · 2026-10-05T02:34:15.668Z Question: For the three pictured frames, should each switch between Shots and Vectors, or would you prefer one dedicated Shots player and one Vectors player? I’ll keep the character beside them unless you want that removed too. Answer: Two players: Shots and Vectors ## H7466 · 2026-10-05T02:34:15.676Z do not keep this ## H7484 · 2026-10-05T02:34:47.389Z make sure the corners are rounded like the orbit ## H7509 · 2026-10-05T02:36:33.842Z replace this with ## H7545 · 2026-10-05T02:37:58.063Z i want to pause on any changes - make the updates from here to here but dont change anything about the larger home layout ## H7598 · 2026-10-05T02:40:50.674Z Question: Two dates in the reference differ from your published résumé: Traversal shows 2024–2026 versus 2025–2026, and JLL shows 2019–2016 versus 2017–2019. Which years should I use? Answer: Use the reference exactly ## H7652 · 2026-10-05T02:43:10.093Z i want to move the socials to be inline but spread apart from Charlie Weston in the h1 and i want arca / design enger ... to be on two lines as to add vertical height ## H7671 · 2026-10-05T02:44:29.416Z make sure that the corners of the images match the orbit ui ... and make sure that these are custom from the admin portal - i should be able to populate from what is on the photos page or custom upload ## H7684 · 2026-10-05T02:44:57.441Z call this 'shots' and 'morphs' ... have morphs be the default ... and remove the 01 / counter ## H7704 · 2026-10-05T02:45:38.779Z remove the 'photos and generative .... through my lens' line ... and make sure the hyperlinks have a WCAG hover state ## H7939 · 2026-10-05T03:06:35.381Z i want to see 4 ways the arca / design eng can be blocked out better here ## H7958 · 2026-10-05T03:07:21.188Z if helpful, add a line with a dot about being in nyc ## H8036 · 2026-10-05T03:14:14.808Z remove this for now ## H8047 · 2026-10-05T03:14:37.868Z remove 'evolving optics' and move 'pause' to be here ( and use the icon button thats pressed in ) ## H8073 · 2026-10-05T03:16:47.087Z call background "education" swap 2016 and 2019 in JLL make Traversal 2025-2026 ## H8075 · 2026-10-05T03:16:47.115Z change github activity to 'exploration' and make the chart to be inline with the title ## H8138 · 2026-10-05T03:20:03.081Z where shoudl i review the work? ## H8149 · 2026-10-05T03:20:38.217Z remove this from the photo block ## H8151 · 2026-10-05T03:20:38.221Z 404 ## H8253 · 2026-10-05T03:25:05.809Z 1 ## H8270 · 2026-10-05T03:26:07.234Z make sure the height lines up ## H8367 · 2026-10-05T03:33:08.537Z remove the social line here and add the how i work as bullets below the 'accors design ive' paragraph ( but make tis smaller ) ## H8389 · 2026-10-05T03:36:26.170Z and then help square this up so that the headins all line up - this looks bad ## H8391 · 2026-10-05T03:36:26.178Z [Image-only feedback; described in the linked decision record.] ## H8532 · 2026-10-05T03:44:52.209Z add vertical space to this so it breathes - come up with 3-4 words that 'design philopsy' could be but i want it to be a single word that starts with an E ## H8619 · 2026-10-05T03:48:21.717Z imrpove this hover state ## H8632 · 2026-10-05T03:49:10.315Z and for the education ## H8658 · 2026-10-05T03:50:19.223Z use ethos ## H8675 · 2026-10-05T03:51:06.494Z align all of this be follow the column witdth of the paragraphs above ## H8688 · 2026-10-05T03:52:16.446Z apply it to 'ethos' and make those columns ## H8777 · 2026-10-05T03:56:45.787Z make these hexagons not numbers ## H8974 · 2026-10-05T04:05:09.755Z Question: For the public case study, which audience should lead? I’ll keep the full decision and artifact record alongside it either way. Answer: Both equally — balance process lessons with portfolio evidence ## H8976 · 2026-10-05T04:05:09.760Z i want to show the design process and eng tradeoffs # Paper comments ## c-09byj3 · biography this is a clean lock up - i really like this ## c-3q5ttp · company-row-3 this is clean AF ## c-5etvaj · activity-title this is nice consistence ## c-9ochuj · company-row-0 i love how background and experience are framed here ## c-bg0zoc · credentials the color needs to be different here ## c-igk6tl · 01 — Quiet confidence i think this is a very clean lock up ( i'm passionate about' and the 2 column bio below ## c-jjcwba · letter i really like this idea of this being a letter and having a slider below to do some text aniamtion and plays with ## c-k2abob · context i like this lock up a lot ## c-r1sovi · narrative-col-2 i need this to be styled differenlty - h6 like this look ai made