docs: record what is actually outstanding in the blog backlog (2 retire, 3 edit — none publishable as-is) - #8
Open
moshebeeri wants to merge 1 commit into
Open
Conversation
The backlog was never enumerated, only referred to. Measured today against
agent-ceo-website origin/main bdcc18be: 290 source posts vs 369 published, 9
source slugs absent — but 2 are published under RENAMED slugs and 2 are
legitimately future-dated, leaving 5 genuinely overdue.
The useful finding is that "5 posts to publish" is the wrong shape. None of the
5 is publishable as-is, and 2 should never ship:
RETIRE self-hosted-super-agent-ceo-nodes 85-90% already served by four
published posts; its `go install` command cannot work for any reader
(repo private, proxy.golang.org 404)
RETIRE in-cluster-deploy-pipeline-ai-agents near-verbatim duplicate of a post
published 2026-06-06, two days BEFORE this one's own date; also
carries a fabricated DEPENDENCIES snippet
EDIT privacy-first-analytics-cookieless only clean one; flagship section
claims "merged but not shipped" and Plausible is live on both hosts
EDIT observable-loop-org-scale 2 dead relatedPosts; thesis is its
third telling; observer/observed conflicts with governor/operand
EDIT super-agents-headless-cli-fleets blocked on the above; MCP tool
table prints names that do not resolve
Also records the instrument warnings that made this hard to see: docs.agent.ceo
is a catch-all that 302s on every path with a byte-identical body, link
extractors that match only `](/path)` miss absolute same-origin URLs, and
`git fetch origin main` silently fails in agent-ceo-website leaving a stale
cached ref.
Decisions, not certifications — the work each implies still needs review, and
nothing here has reached a served page.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
moshebeeri
added a commit
that referenced
this pull request
Aug 16, 2026
…be published (#9) Founder decision 2026-08-16. Completes the backlog decision record (docs/backlog-decision-2026-08-15.md, PR #8): of the five overdue posts, three are published and these two are retired. WHY A NEW TOP-LEVEL retired/ AND NOT posts/retired/ — this is the whole point. Both automated consumers glob posts/**/*.md RECURSIVELY: .cicd/build.sh:14 git diff --name-only ... -- 'posts/**/*.md' scripts/generate-sitemap.py:91 glob.glob('posts/**/*.md', recursive=True) A posts/retired/ subdirectory would still be matched, and the next converter run would publish them anyway. Retirement has to be structural or it is decoration. Verified after the move: both globs return 0 hits for these two files, and the sitemap post count goes 290 -> 288. Nothing is deleted. Both drafts stay in git and in retired/, each with a RETIRED block recording the date, the decision-maker and the evidence. Reviving one is a git mv; it just cannot happen silently. RETIRED: self-hosted-super-agent-ceo-nodes (2026-05-29) ~85-90% already served by four posts, all verified live 200 on stg with a known-negative 404 in the same run: 2026-05-29-self-hosted-super-agent-ceo, 2026-06-01-super-agent-ceo-self-service-docs, super-agent-ceo-backend-mode, super-agent-ceo-launch. Hard blocker independent of overlap: its install command cannot work for ANY reader. Re-measured today — proxy.golang.org/.../agent-hub/@v/list -> 404 github.com/GenBrainAI/agent-hub (anonymous) -> 404 (private) install.agent.ceo/super-agent-ceo (published) -> 200 The material worth keeping was a section, not a post, and went into the docs. RETIRED: in-cluster-deploy-pipeline-ai-agents (2026-06-08) Near-verbatim duplicate of in-cluster-deploy-cloud-build-api-gke, which is dated 2026-06-06 — TWO DAYS BEFORE this draft's own date — and is live at 200. The published version is strictly better (git sync, component->K8s mapping, smart-deploy skip, rollout timeout, post-deploy smoke tests). It also carries a FABRICATED code sample: `declare -A DEPENDENCIES=(...)` appears 0 times in the real scripts/cluster-deploy.sh, which uses case/string comparison and includes `fullstack` (11 mentions) that the draft omits. Every claim above was re-verified against the live surface or the source before being written down, not carried over from the earlier analysis. Co-authored-by: Marketing Agent <marketing@agent.ceo> Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The backlog has been referred to for weeks but never enumerated. This records the measurement and the decisions.
Measurement (2026-08-15, vs
agent-ceo-websiteorigin/mainbdcc18be)posts/**content/blog/*.mdxRoot cause: the cicd converter has been dormant since 2026-05-31, so a post merged here is a file in a source repo, not a page.
The finding: "5 posts to publish" is the wrong shape
None of the 5 is publishable as-is, and 2 should never ship.
self-hosted-super-agent-ceo-nodesgo installcommand cannot work for any reader (repo private, proxy.golang.org 404)in-cluster-deploy-pipeline-ai-agentsDEPENDENCIESsnippetprivacy-first-analytics-cookielessobservable-loop-org-scale-engineeringrelatedPosts; thesis is its third telling; observer/observed conflicts with the governor/operand definition of recordsuper-agents-headless-cli-agent-fleetsEvery claim was re-verified against the served surface or against code before being written down — Plausible's script tag is present on
agent.ceoandstg.agent.ceo; the duplicate post returns 200; the three dead slugs 404 and exist nowhere in the 369.Editorial issue bigger than any one post
"super-agent" already means four incompatible things in the published corpus — the AI-CEO orchestrator, a pool of generalist runner pods, scale-to-zero sub-intelligences sold as a pricing bundle, and your own machine as an operand. The blocked post would add a fifth. That is a term we rank for; resolving the naming is worth more than shipping the post.
Instrument warnings, recorded so the next audit doesn't repeat them
docs.agent.ceois a catch-all — every path returns 302 with a byte-identical body, including nonsense paths. A 200/302 from that host proves nothing.](/path)misseshttps://agent.ceo/path. That gap made a corpus audit report "zero dead links" while four live CTAs were 404 (agent-ceo-website PR #894).git fetch origin mainsilently fails in agent-ceo-website, leaving a stale cached ref — found 8 commits behind.Open question
Nobody has confirmed whether the converter preserves
relatedPostsfrontmatter, or how it maps source slugs to published slugs. If it drops frontmatter, two of the dead-link items are moot. Worth settling before anyone spends time on them.Note
These are decisions, not certifications. The work each implies still needs review, and nothing here has reached a served page.
🤖 Generated with Claude Code