Every NDJSON event used to commit its own React update and re-parse the
whole partial answer through react-markdown, so long replies grew
quadratically slower. Stream events now land in a frame buffer that
flushes at most once per animation frame, releases answer and thinking
text at a steady pace with a twelve-frame catch-up, and settles
synchronously on completion, failure and abort. Streaming markdown is
split at the last completed block so only the tail is re-parsed each
frame. Applied to both the consultation hook and the rectification chat.
BUG-473
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JUei7K13cYxLHE3Axe4A45
Both gate jobs spent 4-8 minutes (validate) plus 3-5 minutes (publish) on
`git fetch --depth=1` of a 105 MB tree over the WAN, and run 2270 burned
15 minutes on three timed-out attempts. act_runner's hostexecutor discards
the workspace between runs but keeps the host filesystem, so keep a bare
mirror at /root/.cache/jyotisha-mirror.git: take an flock on it (validate,
publish, or overlapping runs may race), clone it once or `fetch --prune`
only when it lacks the requested commit, fetch the exact SHA from local
disk, then point origin back at Gitea for every later step.
The mirror is an accelerator, not a dependency: an unwritable path, a busy
lock, a failed clone/fetch, a non-repository directory, or a stale git lock
file all fall through to the unchanged bounded three-attempt remote fetch,
and the SHA-format, checkout, `git clean -ffdx`, HEAD-equals-SHA, and
clean-tree assertions are untouched. Rehearsed locally against a local
remote for fresh, reused, prune-refresh, unwritable, corrupt, stale-lock,
and lock-contention cases.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VawU7Xfd5jS9wUEXz1XYmS
Twenty-four of the last sixty staging pushes were pure documentation, yet each
one ran (and cancelled) the full gate and image publish. Introduce
deploy/gated-paths.txt as the single source of truth for what must rerun the
gate: every Dockerfile COPY source, the Python package inputs, the workflow and
build-context files, and the repository files frontend/tests read at gate time.
Both triggers of backend-quality-gate.yml now carry that exact list; pushes
that touch none of it neither run the gate nor cancel a running code gate.
Because staging head may then legitimately sit ahead of the last tested SHA,
add deploy/is-docs-only-range.sh: it proves <base> is an ancestor of <head>
and that no changed path matches a gated glob, from local history when it is
available and otherwise from the Gitea compare API (per-commit `files`, parent
walk for ancestry, total_commits cross-checked). The publish dispatch and the
deploy-staging head checks accept an advanced head only when that script
succeeds; diverged, older, or code-bearing heads are still refused. In
deploy-staging the check runs after the gate-attested controller bundle is
extracted so only the tested checker and path list are ever executed; the
manual rollback branch is unchanged.
AGENTS.md §6.3/§6.4 describe the new contract: `.deployment.gitCommit` must
equal the latest staging commit that touched a gated path, not staging head.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VawU7Xfd5jS9wUEXz1XYmS
deploy/reclaim-runner-disk.sh ran `docker builder prune --force --all` at the
start of every gate job, so each publish rebuilt the Dockerfile `npm ci` layer
from scratch behind the registry mirror (7-minute image builds became 47).
Reclaim build cache in tiers instead: prune entries unused for 72h, re-measure
free space, and escalate to `--all` only while still below MINIMUM_FREE_GIB.
The final threshold check and the "only unheld resources" rules are unchanged.
Contract test asserts the aged prune runs first and that `--all` is only
reachable inside the re-measured conditional.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VawU7Xfd5jS9wUEXz1XYmS
Gitea (git.copse.top) is the only CI/CD control plane; GitHub is a
read-only mirror whose Actions are being disabled in repository settings.
- Delete all 11 `.github/workflows/*.yml` (stale copies of the old design).
- Delete unused Gitea manual workflows `ci.yml`, `test.yml`,
`publish-pypi.yml`, `apply-supabase-profile-migrations.yml`
(0-1 historical runs, no remaining target).
- Fold the full `python -m pytest` tree and `tests/run_all.py` into
`release-quality-gate.yml`, which previously only ran the curated
release profile; update the `run_quality_gate.py` comment accordingly.
- Port `reset-staging-account.yml` to Gitea: `runs-on: xiaoxin`, bounded
exact-SHA checkout, `refs/heads/staging` only, `staging-mutation`
concurrency, same email/confirmation/host/port/user/path assertions.
- Repoint frontend workflow tests at `.gitea/workflows/`, drop the
GitHub-only assertions, add coverage for the new reset workflow, and
remove `tests/test_supabase_profile_migration_workflow.py`.
- Update AGENTS.md §6.8, README.md, and deploy/README.md to the current
production/mirror facts and document the staging account reset.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VawU7Xfd5jS9wUEXz1XYmS
Choice legends and spoken collect prompts were hidden after BUG-461, and occupation still blocked canAdopt once dated coverage was done. Restore visible stems and stop polling extra collects when adopt is allowed.
Co-authored-by: Cursor <cursoragent@cursor.com>
eslint-plugin-react-hooks now analyzes the smaller Home, so render-time ref writes and sync effect setState were failing the staging gate.
Co-authored-by: Cursor <cursoragent@cursor.com>
Close the home-page split by moving account/onboarding and rectification glue out of page.tsx without relocating React state, keeping the route static and hook order intact.
Co-authored-by: Cursor <cursoragent@cursor.com>
Reviewed: tsc + consult 295 + rectification 747 + python suites green;
guard functions and window/general instructions verified unchanged.
Conflict: BUG entry renumbered 467→468 (467 taken by report-extraction).
Review fix: test_consultation_birth_accuracy.py added to CORE_PYTEST_TARGETS
(pytest-style file outside the rectification glob would never hit the auto gate).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0155nFCgCHtoA7jhSDGmZmMu
The report writer had no interpretation methodology at all: a local agent
calling the jyotish skill can read the reference library, the report model
could read nothing. It could only restate the bundle.
- frontend/src/lib/report-interpretation-packs/ holds one general pack and
one pack per report theme, distilled from the in-repo reference guides.
They constrain wording and reasoning discipline (term modernisation,
how to talk about relative strength and SAV scores, the reasoning errors
to avoid, the banned phrasings) and never assert a chart fact.
- The general pack rides INSIDE the cached system message so the cached
prefix stays byte-identical across sections; the chapter pack follows it
and summary calls get the general pack only.
- Skill jyotish-personal-report goes to 1.1.0 (1.0.0 deprecated): the
contract now names interpretiveFacts and themeNarrativeSeeds as a
bounded fact layer and states that the knowledge pack is not a fact
source and cannot raise certainty.
- Telemetry records interpretiveFactCount and knowledgePackCharacters as
numbers only; the counter never throws so telemetry cannot break a run.
Tests lock every theme resolving a pack, the 3,000 character budget, a
forbidden-substring scan (paths, module names, vendor names, artefact
names), the byte-stable cache prefix, and that no evidence id or date
appears in the static content.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016P5RoqzmUQEbeC2qjAkeGr
The consultation workflow already returns a functional benefic/malefic
table, a shadbala ranking, SAV scores, the current maha/antardasha,
detected yogas and guided-topic copy. The report extraction layer threw
all of it away, so claim cards could only say "the server closed the
minimum evidence group" and the writer had no conclusions to work from.
- ReportEvidenceBundleV2 gains interpretiveFacts (yogas, functionalRoles,
shadbalaRanking, savScores/savTotal, currentDasha, convergenceDomains)
and themeNarrativeSeeds. Both are required, allow empty, keep .strict(),
are covered by the canonical sort + bundleHash, and fail closed on
dangling refs, duplicate ranks/houses/themes and out-of-bound text.
- Extraction is allowlist-style: closed enums for yoga category and
functional role, safeCelestialName for planets, sign->whole-sign-house
projection for SAV, and a forbidden-token scrub that drops any seed line
naming an external provider or internal route.
- Claim card conclusions and supportingFacts are now deterministic
astrological statements built from those facts; risks become
counterFacts. assertionLevel derivation is unchanged, and a theme with
no seed keeps the old receipt wording with consensus capped down.
- filterReportEvidenceBundleForSection trims seeds and SAV houses to the
chapter's theme while letting the chart-wide interpretive receipts ride
along, so every section can cite them.
Contract snapshot taken from a real local /api/consultation_workflow call
with fictional smoke birth data; the new fixture test locks the shapes
that call actually returns, including the fields that are absent.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016P5RoqzmUQEbeC2qjAkeGr
Keep archive as a PATCH of archived_at rather than DELETE, pass entry_mode through onboarding, and let popstate to the default chat reuse selectSession. Then move send/stop/recovery and session management out of page.tsx so the home surface stays within the batch-two line budget.
Co-authored-by: Cursor <cursoragent@cursor.com>
Natal answers were opening on parameter tables, and rectification turns were one-sentence legal copy. Centralize user-facing strings, keep representative-minute and question-slot red lines, and stop duplicating the opening collect prompt as a second assistant message.
Co-authored-by: Cursor <cursoragent@cursor.com>
Hospital records and adopted rectification times were still fed to the model as not_auto_rectified because the chart request omitted declared_accuracy/time_source and mastra hardcoded the boundary. Map profile truth into the engine request, keep rectified for accepted/confirmed active times only, and leave window/general guards unchanged.
Co-authored-by: Cursor <cursoragent@cursor.com>
page.tsx still owns the chat main chain, but the first product surfaces now
live in their own modules so later splits can land without editing the 4k-line
Home. Source-lock tests follow the moved tokens; the orphan user-data contract
is aligned and added to the quick gate.
Co-authored-by: Cursor <cursoragent@cursor.com>
Local fallbacks were creating fake saves and resurrecting deleted rows.
Pin and archive now live on chat_sessions so they follow the account.
Co-authored-by: Cursor <cursoragent@cursor.com>
Refresh, back, and login return were dropping the open conversation because selection lived only in React state.
Co-authored-by: Cursor <cursoragent@cursor.com>
List GET no longer ships transcripts; consult appends questions after reserve and ignores client history so dual-tab last-write-wins cannot erase messages.
Co-authored-by: Cursor <cursoragent@cursor.com>
Window_scan assertions now match the public from_sign/to_sign contract, the
staging quick gate runs the rectification Python suite, and compare-candidates
rescores when stored policy lags the live engine identity.
Co-authored-by: Cursor <cursoragent@cursor.com>
Day-level events now score Vimshottari/Narayana transition closeness so
nearby candidate minutes can diverge, with gated quality probes and
answer-prior ranking so high-base-rate existence questions stay out.
Co-authored-by: Cursor <cursoragent@cursor.com>
deliveryCapability no longer ties adoption to minute separation or holdout;
those stay on the exact-minute confirmation gate so users can save a range.
Co-authored-by: Cursor <cursoragent@cursor.com>
Spoken collect no longer renders a second visual prompt; choice legends stay screen-reader only and live cards share the assistant inset. Exhaustion collect avoids colliding with the opening question id, and a successful billed turn no longer surfaces run_failed after the exit gate.
Co-authored-by: Cursor <cursoragent@cursor.com>
.personal-report-reader declared --report-paper, --report-rule and
--report-accent on itself. A custom property resolves from the nearest
element that declares it, so an element-local declaration beats :root
regardless of media queries or source order — the dark override added
with the theme was dead the moment it was written.
The inks did follow the theme, because they live on :root. So dark mode
put --color-ink (#f2f0ea) on paper that stayed #f8f5ee: 1.05:1 for
headings and 1.22:1 for body. The report was effectively invisible.
Printing had the same defect from the other side: the print block forces
white backgrounds but never reset the inks, so printing from a dark
screen produced a blank page. It now pins the palette back to light —
paper is always paper.
Every report ink now clears AA on report paper in both themes, and the
new contract test forbids re-declaring the report palette anywhere but
:root, which is the trap that caused this.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0155nFCgCHtoA7jhSDGmZmMu
Jesse_Chen