Presentation layer (HTTP val) that fetches JSON Resume data from linkedin-mdp-api and renders it as a styled HTML page in the browser.

Architecture: This val is a pure UI layer. It contains zero LinkedIn/OAuth logic — all data fetching and normalization lives in linkedin-mdp-api.

---

URL	Description
/	302 redirect to the route configured in DEFAULT env var
/minimal	Serif CSS-only theme (print-friendly)
/tailwind	Card-based theme via Tailwind CDN
/jsonresume	Server-side render using a jsonresume.org theme

Param	Description
?photo	Show basics.image as avatar; absent = no photo. Applies to all routes.
?photo_pos=x:y	Crop anchor for avatar: both values 0–100, default 50:20 (center-top heuristic for portraits). Either part optional: ?photo_pos=:30 sets only Y. Applies to /minimal, /tailwind.
?force	Bypass cache read; always fetch fresh data from upstream (result still saved to cache). Presence-only — value is ignored. Applies to all routes.

Param	Description
?theme=<name>	jsonresume theme package name (default: even). Any npm:jsonresume-theme-<name> is accepted.

Theme	Markdown handling	Notes
even (default)	✅ native (micromark)	Processes every text field internally; do not preprocess
kendall	✅ sentinel post-processing	Works for all fields: {{{triple}}} and {{double}} both handled
flat	✅ sentinel post-processing	Schema compat (url→website, name→company) applied first
stackoverflow	✅ native (nl2br/paragraphs)	Built-in paragraph transform; sentinel skipped
paper	✅ sentinel post-processing	Applied automatically as unknown theme
elegant	✅ sentinel post-processing	Applied automatically as unknown theme

Why flat shows raw Markdown: flat uses {{double}} braces throughout — Mustache/Handlebars escapes the output. Pre-rendered HTML would appear as literal <p> tags. flat does not support rich text.

Sentinel post-processing (kendall projects.description): Since {{description}} is double-braced in the kendall template, we substitute a unique sentinel string before rendering, then replace the escaped sentinel in the output HTML with the markdown-rendered HTML. This avoids false replacements that would occur with naive text matching.

Themes with exotic runtime dependencies (native addons, missing peer packages like @resume/core) may fail — the error page shows the exact message.

---

Header	Description
X-Cache-Age	Seconds since the last successful cache write. -1 = no cache entry exists yet. 0 = data just fetched fresh.

---

Key	Required	Default	Description
DATA_API_URL	—	hardcoded linkedin-mdp-api endpoint	Base URL of the data API (no trailing slash)
DEFAULT	—	minimal	Default route for / redirect. Can include query string, e.g. tailwind?photo

DEFAULT warning in logs ("DEFAULT" is not set) is harmless — the val falls back to minimal.

---

Resume data is cached in Val Town Blob Storage for 24 hours.

Detail	Value
Blob key	linkedin_resume_cache
Format	{ fetchedAt: number (ms), data: JsonResume }
TTL	24 hours
?force	Skips cache read only; result is still written to cache
Stale fallback	If upstream returns 429/5xx/network error and a stale entry exists (any age), stale data is served silently

---

Browser → /minimal (or /tailwind, /jsonresume)
           └─ Blob cache (24h TTL)
                └─ fetch DATA_API_URL/jsonresume   [on cache miss or ?force]
                     └─ linkedin-mdp-api
                          └─ LinkedIn DMA API (OAuth)

linkedin-mdp-api applies all data normalization:

• Markdown-compatible paragraph/bullet formatting
• photo extraction from RICH_MEDIA log
• JSON Resume schema v1.0.0 compliance

linkedin-resume-ui applies presentation-only transformations:

• md() — renders Markdown to HTML via npm:marked (GFM); used in /minimal and /tailwind
• mdStr() — same, for pre-processing fields before passing to /jsonresume themes
• preprocessAll() — applies mdStr() to all text fields; used when ?preprocess=html
• patchedKendallRender() — kendall-specific: sentinel post-processing for projects.description
• flatSkills() — expands { name: "Skills", keywords: [...] } into chips
• dedupDegree() — removes "X, X" duplicated degree strings
• dateRange() — formats ISO dates as YYYY-MM – YYYY-MM or YYYY-MM – Present
• label normalization — \n-separated role titles → · separator

---

Photos are rendered as circular avatars (96×96 px) using a container div + inner <img>:

• Proportional scaling: the <img> fills its container via object-fit: cover. Source aspect ratio is always preserved — no distortion.
• Crop anchor: object-position is controlled by ?photo_pos=x:y. Default is 50% 20% (center horizontally, slightly above center vertically — a heuristic that works well for LinkedIn-style portrait photos where the face tends to be in the upper third).
• /jsonresume: basics.image is stripped from the payload when ?photo is absent, because themes render the photo unconditionally if the field is set.

---

Situation	Behaviour
LinkedIn API 429/5xx + stale cache	Serve stale cache silently
LinkedIn API 429/5xx + no cache	Pass through response as-is
Network error + stale cache	Serve stale cache silently
Network error + no cache	502 + error page with DATA_API_URL hint
API returns non-JSON	502 + error page
basics.name missing	502 + error page
Unknown route	404 + error page listing valid routes
Theme has no render() export	500 + error page with list of working alternatives
Theme import/render throws	500 + error page with exact error message and Deno hint

---

Version	Key changes
v1–v3	Initial themes, token passing, basic structure
v4	Skills flatting, degree dedup, prose(), label normalization
v5	DEFAULT redirect, ?photo, jsonresume server-side render
v6	API error passthrough, basics.name validation, object-fit: cover
v7–v10	npm: import fix, theme compatibility, blob cache (24h TTL)
v11–v12	Stale-while-revalidate, X-Cache-Age, ?force, ?photo_pos=x:y
v13	prose() → marked (full GFM), proportional photo, ?preprocess=html, per-theme markdown map
v14	kendall CSS override (#photo img), sentinel post-processing for projects.description
v15	?force changed to presence-only (was =1); README cleanup and version sync

---

• Theme selector UI — a nav bar / dropdown listing available routes and themes. Input field should be editable (free text) to support unofficial jsonresume-theme-* npm packages beyond the official list. Official theme list source: npmjs.com search: jsonresume-theme

• Banner support — linkedin-mdp-api exposes a banner/background image URL in the raw data. JSON Resume schema does not have a standard banner field. Investigate: (a) whether any jsonresume themes support it via a non-standard field; (b) adding a banner strip to /minimal//tailwind.

• basics.image proxy — LinkedIn photo URLs may expire or require auth; consider proxying through the val to avoid broken images.

Approaches evaluated (priority: minimum code + fast):

Approach	Effort	Speed	Quality	Notes
print.css polish for /minimal	Minimal	Instant (browser Ctrl+P)	Good	/minimal already has @media print skeleton; needs page-break rules and margin tuning. Recommended first step.
Puppeteer /pdf route	~10 lines	~1s	Excellent	Requires npm:puppeteer or Val Town @browserless; cold-start latency
npm:marked + Pandoc WASM client-side DOCX	~20 lines JS	5–20s (WASM load)	Good	Client-side, no server cost; Markdown is the source format

Rejected approaches:

• RenderCV — Python/Pydantic/LaTeX CLI only; no WASM, no serverless path.
• Pandoc server-side — Haskell binary; cannot run in Deno/Val Town.
• OpenResume — React SPA with proprietary JSON schema; no public render API.
• HackMyResume/FluentCV — abandoned (last commit 2018).

• <meta og:…> tags for social sharing preview.
• Dark mode via @media (prefers-color-scheme: dark).
• ?lang= param to filter multi-language data (dependent on API support).
• Extend sentinel post-processing to flat — verify that flat's template renders a projects section, then apply the same per-field sentinel approach used for kendall.

---

Tool	Format	Render	Status
JSON Resume	JSON	30+ npm themes	✅ In use
RenderCV	YAML	LaTeX/Typst → PDF CLI	CLI only, no serverless
OpenResume	Proprietary JSON	React SPA → PDF	No public API
Manfred MAC	MAC JSON	getmanfred.com SaaS	No open-source renderer
HackMyResume	JSON Resume / FRESH	HTML/PDF/DOCX CLI	Abandoned since 2018
Pandoc	Markdown	DOCX/HTML/PDF	Useful as Markdown→DOCX (WASM, client-side)

---

LinkedIn DMA API has a per-day rate limit. Once exhausted, all endpoints return 429 until midnight UTC. The stale cache fallback means the UI continues serving the last known data silently.

Themes control which sections they render. Some (e.g. flat) do not include a projects section in their template — this cannot be fixed without forking the theme.

Themes imported via npm:jsonresume-theme-<name> work if they only depend on pure-JS packages. Themes with native addons or missing peer packages (e.g. @resume/core) will fail — the error page shows the exact message.

Val Town emits WARNING: "DEFAULT" is not set on every request. Cosmetic only — set the env var to silence it.

Profile edits may take hours or up to a day to appear. LinkedIn DMA API limitation.

---

• foxdark410/linkedin-mdp-api — data layer: fetches LinkedIn DMA snapshot, exports JSON Resume, MAC, RenderCV, Markdown, plain text.

---

linkedin-resume-ui/
└── main.ts   # Single HTTP handler — all themes, routing, helpers in one file

All logic intentionally lives in one file to keep the val self-contained. If the file grows beyond ~1000 lines, consider splitting into:

• themes/minimal.ts, themes/tailwind.ts, themes/jsonresume.ts
• utils.ts (md, esc, dateRange, flatSkills, dedupDegree, photoCSS)
• main.ts (router + cache + error handler)

• npm packages: import X from "npm:package-name"
• Val Town std lib: import { blob } from "https://esm.town/v/std/blob"

When the LinkedIn API limit is hit (429), test by temporarily hardcoding a fixture object in main.ts and bypassing the fetch call:

// Temporary fixture for local dev / rate-limit debugging
const json = { basics: { name: "Test User", label: "Engineer" }, work: [] };