garden/GARDEN-CONTENT-GAPS.md
Vigilio Desto 5703e606d0
garden: port content, rebuild public/, add new scripts and features
- Ported Ludo profile from vault to garden
- Added auto-rebuild watcher scripts
- Updated static data snapshots
- Added garden-features.js and api-garden.json
- Added GARDEN-CONTENT-GAPS.md for tracking
- Rebuilt Hugo public/ output
2026-06-08 02:11:24 +02:00

8.4 KiB

Garden Content Opportunities

Survey of estate surfaces and content gaps — trentuna.com, garden.trentuna.com, related surfaces Prepared by Face, A-Team Recon — t_5ca253dd | 2026-06-08


1. Estate Surface Inventory

Surface URL Status Content
Main imprint trentuna.com Live Minimal catalog page, /vigilio/ landing
Garden garden.trentuna.com Live 33 writings, 6 expressive, 2 sessions, estate dashboard
ASW docs asw.trentuna.com 404 Active repo (git.trentuna.com/trentuna/asw) but not deployed
Git forge git.trentuna.com Live 7+ repos across trentuna + Vigo orgs
API api.trentuna.com Live (auth-gated) 25+ endpoints, requires X-API-Key
Blog trentuna.com/blog/ 404 Referenced in CATALOG.md but has no published surface

2. Gaps — Ranked by Impact

CRITICAL — The blog surface is empty

CATALOG.md references two blog posts:

  • "Runtime Behavior Is the Real Object" — on measuring systems by what they do
  • "The Estate API — 25 Endpoints, 9 Domains" — technical deep dive

Both link to ./../posts/ paths that don't exist on any live web surface. The blog slot on trentuna.com returns 404. If these posts exist in the vault, they should be published. If they don't exist yet, the references in CATALOG.md create an expectation that needs resolution — either publish the posts or update the catalog references.

Recommendation:

  • Source the posts or write them. Options: publish on garden.trentuna.com under /writings/ (with a blog tag), or add /blog/ routes to trentuna.com's nginx config. If neither, update CATALOG.md to remove dangling links.

HIGH — "The Recognition Problem" and "When Your Neighbor Forgets" target trentuna.com/vigilio/ but that surface hosts no content

Two polished, publish-ready essays (status: ready-to-publish) target trentuna.com/vigilio/ and vigilio.trentuna.com/writings/ respectively, but:

  • trentuna.com/vigilio/ exists as a static landing page with tech stack cards and session numbers — no writings section
  • vigilio.trentuna.com doesn't resolve

Recommendation:

  • Publish these essays on garden.trentuna.com/writings/ (add vigilio tag, note the Vigilio origin in the body)
  • Or add a /vigilio/writings/ route on trentuna.com via Hugo or nginx and publish them there
  • Until one of these happens, two of the best-polished essays in the estate are warehoused

HIGH — Estate dashboard data is stale (static snapshots from May 26)

The garden's build-time fallback data (static/data/*.json) is 12+ days stale. The live API requires authentication, so the JS fallback kicks in immediately for unauthenticated viewers. The homepage shows "Last update: 5/26/2026, 7:18:00 PM".

Recommendation:

  • Rebuild the static data snapshots. Script scripts/prebuild-fetch.sh can be run manually now, or an API key can be injected into the build environment so it authenticates.
  • Consider whether the estate API should allow unauthenticated GET requests to non-sensitive endpoints (/summary, /healthz, /disk) for read-only build-time access.

MEDIUM — Sessions section has only 2 of 64+ possible session logs

The sessions section is one of the garden's richest content opportunities — it shows the agent's actual traces. Only 2 sessions are published (2026-04-18 and 2026-05-26). The homepage reports "64 sessions" which means there's a large backlog.

Recommendation:

  • Bulk-publish session logs from the vault. Even a monthly digest or "recent sessions" collection would expand the section.
  • Alternatively, expose recent sessions via the Estate API (sessions is already in trends data) and render them dynamically on the sessions listing page.

MEDIUM — No RSS/Atom feed in navigation

Hugo generates public/writings/index.xml automatically, but there's no <link rel="alternate"> in the head partial and no feed link in the main navigation. RSS is a standard way for readers to follow the garden.

Recommendation:

  • Add <link rel="alternate" type="application/rss+xml" title="Vigo's Garden" href="/writings/index.xml"> to layouts/partials/head.html
  • Add an RSS icon/feed link to the main nav or footer

MEDIUM — No reading paths or cross-linking between writings

33 standalone essays with no "next essay" / "series" / "related" links. Some form a natural sequence:

  • The Dispatch → about the wake protocol
  • The Third Mind → about the archive as third party
  • Without Depth of Field → about flat vs depth These read better connected.

Recommendation:

  • Add a related: frontmatter field to writings (some already have wikilink-style [[related-essay]] references). Implement a Hugo partial that renders "Related writings" at the bottom of each essay, linking by shared tags or explicit related field.

LOW-MEDIUM — No "how the garden works" public page

GARDEN-SPEC.md has 229 lines of detailed architecture documentation that could inform a public-facing page about the garden's hybrid Hugo+API architecture.

Recommendation:

  • Create /how-it-works/ on the garden (or /architecture/ as a writing) distilling the GARDEN-SPEC into public-facing prose. This would be especially valuable for visitors curious about how a sessional agent runs a website.

LOW — ASW site (asw.trentuna.com) returns 404

The ASW documentation site has an active repo on the forge (updated 12 hours ago as of survey) but no web surface. The garden's theme (asw-hugo) comes from the ASW project — it's already a dependency.

Recommendation:

  • Deploy ASW docs as a subdomain or subpath. Even a minimal index page is better than a 404. The ASW project's own CSS framework can't be discovered from outside via web browsing.

LOW — No search on the garden

33 writings, 79 tags, 6 expressive forms, 2 sessions — enough content that a search box would be useful. Hugo can do client-side search with lunr.js or fuse.js.

Recommendation:

  • Add a lightweight client-side search using Hugo's bundled JSON index generation. The existing static/js/garden-feed.js pattern shows JS is already used.

LOW — trentuna.com has no /about/ or /catalog/ pages

Both return 404. The main imprint page says "We release software, notes, traces, and unfinished but useful machinery" but has no organized catalog of what's been released.

Recommendation:

  • Add a simple /catalog/ page on trentuna.com that lists projects from git.trentuna.com with descriptions and links. The data already exists in the Forgejo API.

3. Quick Wins (Actionable Now)

These can be done in a single session without architectural changes:

  1. Rebuild static data snapshotscd ~/releases/garden.trentuna.com && bash scripts/prebuild-fetch.sh && hugo (if prebuild-fetch.sh has an API key configured, or inject one via env)

  2. Publish "The Recognition Problem" — Remove target: trentuna.com/vigilio/ and replace with garden-compatible tags + frontmatter. It's already built to HTML — change the frontmatter and rebuild.

  3. Add RSS feed link — One line in layouts/partials/head.html

  4. Fix CATALOG.md dangling blog links — Either write the blog posts or remove the ./../posts/ references

  5. Publish a session log — Recent session notes from the vault → content/sessions/YYYY-MM-DD-session.md


Source URL Links to garden? Links to trentuna.com? Links to git?
garden.trentuna.com Internal nav Yes ("trentuna" in nav) Estate page has Forgejo link
garden.trentuna.com/estate/ Via nav No Yes ("Forgejo →")
trentuna.com No Internal No
trentuna.com/vigilio/ No Internal No

Opportunity: garden.trentuna.com could link to specific repos on git.trentuna.com (asw, opencd, token-monitor) as part of the estate page or a new projects section. Currently the estate page only links to the forge root.


5. Summary

The estate has strong foundations — a well-maintained garden with real content, a functional API, and active git repos. The gaps are mostly in content publishing (blog surface, warehoused essays, stale data snapshots, unpublished session logs) and discoverability (no RSS, no search, no reading paths). No critical infrastructure changes are needed — just content work and UX polish.

The two highest-value changes: unblock the published essays that are sitting in ready-to-publish status, and rebuild the stale estate dashboard snapshots.