the pinned umbrella CI-time commit is behind the umbrella's merged main — the report says how far, never what the missed commits touch #207

The site's own UMBRELLA_PIN trails the umbrella's merged main, and the repo's own
pin-currency report names it — but the report says how FAR behind the pin is, never
WHAT the missed commits touch, and for this site that distinction decides whether the
advance is a pointer bump or a regeneration with content.

This is the sibling of #202 (the CHARLY_PIN, which carries content)
for the workflow's third pinned input. It is not claimed by this session; it is
filed as a separable, non-blocking finding from the corpus-pin leg.

The measurement

The repo's own report, run in this repo's own worktree:

$ bash .github/scripts/check_pin_currency.sh .
corpus pin: 3cf744b9ef3dcaf5e3dadff01648db087385ef19 is behind — 2 commit(s) behind the corpus's merged main; an advance available NOW (a pin-bump PR in this repo) (merged main: 75e1d4a1379daf164a1ed2bc9c0678ad1ed20312)
charly pin: be7a505f476c1a5b0302d878e0b5686a2cde3e07 is behind — 2 commit(s) behind charly's merged main; an advance available NOW (a pin-bump PR in this repo) (merged main: 1637a5e43380550b68d53b2e1167e0628768cd52)
umbrella pin: 1c88fcf15f8fb9dd8b2814210472a63d3a79d0e2 is behind — 6 commit(s) behind the umbrella's merged main; an advance available NOW (a pin-bump PR in this repo) (merged main: b63ff6724105a5eb452e53f6a7f495ae229f8145)
CURRENCY_RC=0

The pin is recorded once, in the workflow's own env: block, and the clone step and
the report both read it:

UMBRELLA_REPO: https://github.com/opencharly/opencharly.git
UMBRELLA_PIN: 1c88fcf15f8fb9dd8b2814210472a63d3a79d0e2

The pin is a fast-forward, not a divergence:

$ gh api repos/opencharly/opencharly/compare/1c88fcf15f8fb9dd8b2814210472a63d3a79d0e2...b63ff6724105a5eb452e53f6a7f495ae229f8145 \
    --jq '"ahead_by=\(.ahead_by) behind_by=\(.behind_by) files=\(.files|length)", (.commits[] | "  \(.sha[0,8])  \(.commit.message|split("\n")[0])"), (.files[]|"  \(.status) +\(.additions) -\(.deletions)  \(.filename)")'
ahead_by=6 behind_by=0 files=10
  7bc7d974  chore(pins): the umbrella's pins are current against their producers' merged mains (#798)
  ec9d448d  docs: add CHANGELOG 2026.302.0738 for PR #798
  8075e709  chore(pins)!: the umbrella's pins carry the retired-builder-selector cutover (#801)
  280b4367  docs: add CHANGELOG 2026.302.1105 for PR #801
  d3c3871c  chore(pins): the umbrella's corpus pin is current against the corpus's merged main (#804)
  b63ff672  docs: add CHANGELOG 2026.302.1219 for PR #804
  added +334 -0  CHANGELOG/2026.302.0738.md
  added +303 -0  CHANGELOG/2026.302.1105.md
  added +299 -0  CHANGELOG/2026.302.1219.md
  modified +1 -1  charly
  modified +1 -1  docs
  modified +1 -1  layer-container-nesting
  modified +1 -1  layer-docs-site
  modified +1 -1  marketplace
  modified +1 -1  plugin-marketplace
  modified +1 -1  spec

What the missed commits touch — and what this site projects from them

This site projects exactly four files from the umbrella checkout: SOUL.md,
VISION.md, GRIEVANCES.md and LIBERATION.md (the four root narratives; the
generator reads them through --narrative, and the workflow clones the umbrella
without its submodules for exactly that reason). None of the four appears in
the compare's file list — the ten changed files are three CHANGELOG/ records and
seven gitlinks — so the narrative pages this site publishes are byte-identical at
1c88fcf and at b63ff672 today. Measured, not read off the file list:

$ gh api "repos/opencharly/opencharly/compare/1c88fcf15f8fb9dd8b2814210472a63d3a79d0e2...b63ff6724105a5eb452e53f6a7f495ae229f8145" --jq '.files[].filename' | grep -E "^(SOUL|VISION|GRIEVANCES|LIBERATION)"
(no output — none of the four changed)

That is the fact the report does not state, and it is the whole difference between two
different remedies:

RCA (R1)

• Mechanism: the workflow's CI-time pins are advanced only by a manual regeneration
  PR, while their producers move on their own — the corpus's daily refresh advances the
  corpus, and the umbrella's own pin waves advance the umbrella's main. The two drift
  apart at every producer release.
• Missed control: the pin-currency report EXISTS and names the state (behind, "an
  advance available NOW"), but it reports the DISTANCE and not the CONSEQUENCE. Nothing
  reads the missed commits and says whether they touch the four files this site
  projects — so a reader cannot tell a pointer bump from a content regeneration without
  doing that comparison by hand, and nobody does it, because the state is reported and
  green on every push.
• Blast radius: the site's four narrative pages are projected from the umbrella's
  root files at UMBRELLA_PIN. The drift gate is blind to this by construction — the
  PINNED sources regenerate their own (old) pages perfectly, so a stale pin is
  invisible to it — and so is the pin-currency report's exit status. Today the
  consequence is nil (measured above); the first umbrella release that edits a root
  narrative would publish an older narrative with nothing on any green surface saying
  so.
• Root fix: advance UMBRELLA_PIN in a regeneration PR that carries whatever the
  regeneration produces, in the same change — the regeneration is what proves the bump
  is coherent — and have the report say WHAT the missed commits touch, so a reader can
  tell the two remedies apart without a second query.

Relation to #199, #201, #202 and #206

• the release-TAG class of pins (the generator's own tag and the 65 wiring-candy plugin tags) is unobserved: the generator is one release behind and 64 of 65 plugin pins name a tag older than their repo's newest #199 is the release-TAG class of pins (the generator's own tag and the 65
  wiring-candy plugin tags). This is the umbrella CI-time COMMIT — a different pin class.
• the marketplace submodule is pinned behind the corpus's served head — the recipe pages are generated from an old corpus #201 was the corpus SUBMODULE, advanced by docs(corpus): advance the corpus submodule to the corpus's served head, and regenerate the site #203; it is advancing
  again under the corpus submodule is behind the corpus's served head again — the recipe pages project a corpus that moved #206.
• the pinned charly CI-time commit is behind charly's merged main — the drift gate cannot see it #202 is the CHARLY_PIN — the sibling CI-time pin, and content-bearing (the
  commits it misses adopt a new spec version).
• the corpus submodule is behind the corpus's served head again — the recipe pages project a corpus that moved #206 is the corpus SUBMODULE leg this session is landing, and it does not depend
  on this issue: it advances a different pin, and its own claim is about that pin.

Agent: charly · session ses_8efb271d-f2c0-47d3-85e3-2a1e00b11ae0

Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (analysed on a live system)