Skip to content

feat(uipath-maestro-case,uipath-planner): SDD schema conformance — combined merge PR (MST-15966) - #3691

Closed
roberts-cliff wants to merge 63 commits into
mainfrom
feat/sdd-schema-mst-15966
Closed

roberts-cliff wants to merge 63 commits into
mainfrom
feat/sdd-schema-mst-15966

Conversation

@roberts-cliff

@roberts-cliff roberts-cliff commented Oct 1, 2026 •

Copy link
Copy Markdown
Contributor

MST-15966. The whole SDD-schema change in one PR, so it lands in one merge. Review it on the stack; merge this. This branch always equals the tip of the stack (#3717), so the diff here is the stack's combined diff. Don't merge the stack PRs one at a time; close them once this merges.

Lands with UiPath/cli#4743 (the CLI rollup). Merge the two close together. Two CLI PRs refuse SDDs that used to pass, and the repair loop that fixes such a document is in this PR. If the CLI lands alone, a planner-written SDD can fail its build with no way to repair it.

Order PR What
1 #3661 S1: degraded SDD corpus, manifest guard, E4 gate check, checklist-coverage record (2/8/16 at cli 4e203e449)
2 #3662 S2: every committed case SDD parses clean; corpus gate in CI
3 #3663 E0: eval preflight checks skill verbs against the build under test
4 #3664 E1a: case stops and hands back when the CLI refuses the SDD
5 #3665 S3a: planner gates the ready flip on sdd validate and discloses repairs (E1b)
6 #3666 S3c: changes to an existing sdd.md route to the planner. Activation gate: case 98.0%, planner 98.9% (85% floor)
7 #3667 S3b: planner runs sdd format on every handoff
8 #3668 E2: sdd format keeps every line a person wrote
9 #3669 E3: convert builds from the model JSON, including the unreadable-sidecar case
10 #3686 S3d: case formats a bare SDD before resolve, so convert reads the sidecar (B7: markdown source 43/43 → sidecar 41/41)
11 #3671 Ledger fix: never "correct" a field sdd resolve wrote
12 #3687 CI: run evals against a UiPath/cli air-gapped archive (cli_archive_run_id), with a manifest-exact audit
13 #3690 A free-form SDD goes to the planner to normalize; the original is kept as sdd.original.md
14 #3700 Say when a case can't run before debug: a runnability check in Phase 6, and "Runs as designed" in the planner's PDD-path summary
15 #3712 Find an unfinished connector rule by its empty slots, not the "placeholder" string (skills half of cli#4589); what the skill writes stays as it is until the CLI ships
16 #3717 S4: sdd validate is the planner's conformance gate; the 35-item walk survives only as a fallback when the CLI has no sdd command. Walk vs no walk: $1.98 → $1.44 mean, 13.7 → 8.6 min, same quality on the measured items. Not ready: the coverage record measures 26 of the checklist's 35 items, and the walk is the only check on the other 9 (see #3717)

Evidence

  • B7 full suite, A/B on cli build 6. Source on readable converts went from markdown 43/43 to sidecar 41/41. On the 67 shared tasks, pass/fail/error was A 63/3/1 and B 61/4/2.
  • CI on archive 36882965827 (cli 52ad1e565):
    • Run 36887671442: 4/4 at 1.0.
    • Run 36912988475: 5/5 at 1.0, including freeform_sdd_normalized and init-validate.

Still owed

🤖 Generated with Claude Code

roberts-cliff and others added 8 commits September 30, 2026 15:47
…15966)

Five fixture SDDs, each an existing fixture with one defect injected
from the measured defect inventory: a table row that loses its leading
pipe, curly quotes in one =js: cell, a renamed Case Variables heading,
a trailing-(annotation) name collision, and an empty document.
manifest.json names every defect, its line, its seed blob, and what one
dev build did with it. That record is not the assertion: the future
degraded_sdd_conformance eval (E4) asserts the gate names each line.

Agent-authored, so it proves the gate catches the measured classes and
nothing past them.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…fest

E4 asserts the gate names each defect's recorded line, so the manifest
has to stay true to the fixtures. Runs in the required maestro-case
checker job: every fixture listed, each recorded line holds the degraded
text and the named defect, and each fixture is its seed plus one line
while the seed is unchanged.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…d corpus

A row without its leading pipe is now read, not rejected (CLI stack A),
so pipe-dropped-row expects read-as-seed; the other four expect the gate
to reject naming their line or the document. Records stack A's results.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…ure's expect

Runs `uip maestro case sdd validate` on every fixture in the degraded
corpus manifest: reject-naming-line exits non-zero naming the recorded
line, reject-naming-document exits non-zero, read-as-seed exits zero and
parses to exactly its seed's Model. Skips while the installed CLI has no
`sdd validate`, like the corpus-clean check. Records what cli 367bfdb69's
gate returned for each fixture.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…rning at line 312

Cliff's call: a row without outer pipes stays readable and `sdd validate`
reports it as the non-fatal SDD_ROW_WITHOUT_OUTER_PIPES. E4 now asserts
that warning, code and line, alongside the read-as-seed model equality.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…the CLI enforces

checklist_coverage.py seeds one violation per planner checklist item into
sla_from_sdd and classifies what the installed CLI does: validate,
validate-warns, convert (refusal or a new Unresolved item, compared by
kind and location), or silent tagged by whether the parsed Model changes.
Rows that convert also record what plain `case validate` raises on the
converted plan, which is after the handoff and so is not gate coverage.

On cli 367bfdb69, checklist_coverage.json: validate catches 0 of 26,
convert reports 8, 18 are silent (10 of them change the Model). The
checklist (skills S4) cannot be deleted until every row reads validate.
Mutations regenerate from the seed each run; one that stops applying is
reported, not skipped.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…date now catches items 7 and 20

cli#4678 (A7) turns a literal \n and an untyped recipient into parse errors
(SDD_LITERAL_NEWLINE, SDD_RECIPIENT_UNTYPED). Tally: validate 2, convert 8,
silent 16 of 26. The previous record (367bfdb69: 0/8/18) is in git history.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… — unchanged (2/8/16)

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 1, 2026 •

Copy link
Copy Markdown
Contributor

Claude finished @roberts-cliff's task in 5m 26s —— View job


Coder-eval task lint (advisory)

  • Read lint rubric
  • Identify changed task YAMLs
  • Check PR body for evidence of passing run
  • Apply rubric to each changed file
  • Within-PR duplicate check
  • Post results

Coder-eval task lint (advisory)

4 task YAMLs changed; verdicts: 0 Critical, 0 High, 0 Medium, 0 Low, 4 OK.

Rubric: .claude/commands/lint-task.md. This check is advisory and never blocks merge.

Evidence of passing run

✅ Author confirms passing run: "CI on archive 36882965827 (cli 52ad1e565): Run 36887671442: 4/4 at 1.0. Run 36912988475: 5/5 at 1.0, including freeform_sdd_normalized and init-validate." B7 full suite A/B also claimed (67 shared tasks, A 63/3/1, B 61/4/2).

Per-task lint

tests/tasks/uipath-maestro-case/freeform_sdd_normalized/freeform_sdd_normalized.yaml — verdict: OK

Well-structured task. Prompt supplies a ground-truth anchor (the SDD text the criteria assert against). Criteria mix skill_triggered, run_command with content assertions (original preserved, receipt present, sdd validate exit 0, design names kept), command_executed, and run_command existence check. No gameability — the planner trigger, byte-exact original check, and CLI validation can't be faked.

tests/tasks/uipath-maestro-case/gate_refusal_hands_back/gate_refusal_hands_back.yaml — verdict: OK

Minimal goal-oriented prompt ("Build it"). Strong negative-evidence criteria: sha256 byte-identity on sdd.md (line 61), no repaired copy (line 69), no hand-authored caseplan (line 77), plus command_executed proving the refusal was reached (line 54) and agent_judge on the handback reply. The agent_judge is backed by deterministic side-effect checks, not the sole grading signal.

tests/tasks/uipath-planner/case/case_correction_leaves_warned_rows/correction_leaves_warned_rows.yaml — verdict: OK

Excellent minimal prompt ("change the SLA from 5 to 4 days, nothing else"). The check_only_requested_change.py custom checker (line 43) validates non-table lines byte-identical and table cells identical except the SLA — strong anti-gameability. Advisory sdd format criterion (line 51, pass_threshold: 0.0) correctly documents the command may not be available. skill_triggered and sdd validate provide cross-checks.

tests/tasks/uipath-planner/case/case_finalize_validate_gate/finalize_validate_gate.yaml — verdict: OK

Tests a specific curly-quote repair scenario. Criteria cover the full surface: file_exists (basic gate, line 62), command_executed for validate (line 70), run_command validate exit 0 (line 77), guard survival with straight quotes (line 85), draft byte-identity via sha256 (line 93), handoff status (line 101), no caseplan (line 109), and agent_judge for disclosed repair (line 116). Prompt is similar in shape to finalize_picker_pairing but tests a materially different defect (quote normalization vs wait-for-user exposure).

Within-PR duplicates

No duplicate clusters detected. The four tasks cover distinct SDD-schema paths: (1) free-form normalization, (2) refusal handback on malformed heading, (3) surgical edit of a warned-but-valid document, (4) curly-quote repair gated by sdd validate.

CLI verb catalog note

Could not run scripts/check-cli-verbs.py (permission). Manual catalog check: maestro case sdd validate and maestro case sdd format are not in assets/uip-catalog-snapshot.json (generated from CLI 1.204.0-dev.8852). These are new SDD-stack verbs shipping with the CLI build this PR tests against (cli_archive_run_id). The passing CI runs on archive 36882965827 confirm the verbs exist in the target build. Consider regenerating the catalog snapshot once the CLI ships these verbs to avoid future checker noise.

Conclusion

✅ All changed tasks pass the rubric. Evidence of passing run confirmed.

@github-actions

github-actions Bot commented Oct 1, 2026 •

Copy link
Copy Markdown
Contributor

Claude finished @roberts-cliff's task in 2m 38s —— View job


PR Review: SDD Schema Conformance (MST-15966)

Reviewing PR — agents analyzing skill files, tests, and CI changes in parallel...

  • Read review criteria and project conventions
  • Read full diff and PR conversation
  • Review changed skill files (SKILL.md, references)
  • Review CI/workflow changes
  • Review test infrastructure (Dockerfile, scripts)
  • Review test tasks and fixtures
  • Check cross-skill references, links, secrets
  • Post final review

Early finding: Personal path /Users/cliff/src/skills-wt-degraded/... in manifest.json:272,277 — these are CLI error messages captured in test fixtures.

roberts-cliff and others added 2 commits October 1, 2026 15:24
…ed CLI messages

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…-15966)

S2 of the SDD-schema stack. Parses each case fixture SDD through the
installed uip and requires Success, at least one stage, and no parse
notes. Scoped to tests/tasks/uipath-maestro-case: a tests/tasks-wide glob
would collect planner documents that are correctly refused as case SDDs.
The degraded corpus is excluded wholesale. A refused document gets a
recorded verdict, never an edit to make it pass.

Runs in its own advisory workflow on @uipath/cli@latest, which has no
`sdd` command yet: the test skips and says so, then checks by itself once
one ships. Not required, since an npm outage or bad release would block
every PR.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
roberts-cliff and others added 27 commits October 2, 2026 12:24
…finalization skips the layers guide before the first Read

Checklist item 6 asked execute-connector-activity for **Resolved Resource:**,
which the template's Connector Task Detail does not have; the gate built from
it (cli#4766) fired SDD_TASK_BLOCK_INCOMPLETE 86 times across 48 S4 runs and
drove 15-28 extra repair ops on the two dearest tasks. cli 54d04b951 drops the
requirement; the checklist now matches the template.

The finalization read budget sat ten lines below "read all three, in
parallel", so 8 of 10 finalize runs read the layers guide anyway. SKILL.md is
loaded before any Read, so the exception now lives there.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…nd sees

The coverage record now seeds every checklist item (cli 54d04b951): validate
20, validate-warns 2, convert 7, excluded 1, silent 3 (17-19: data closure),
needs-draft 4 (30-33: draft parity, which no CLI command can read). Items 28
and 35 are caught, by neighbouring rules (SDD_RULE_NOT_FOR_SLOT,
SDD_WAIT_FOR_USER_UNPAIRED).

The planner checks those seven inline on the written file — 17-19 always,
30-33 when finalizing a draft — each stated in full in the lane, no
subagent, no checklist read. Everything else stays sdd validate's; the full
walk remains only the no-sdd fallback.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… Value, as the template does

The template's Timer Task Detail has no **Duration:**; the gate built from the
checklist fired SDD_TASK_BLOCK_INCOMPLETE "lacks **Duration:**" 4 times in two
S4 runs on cli 36b85f8e4. Same class as the Resolved Resource fix.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… 13 moves to the gate

SDD_SLA_REFERENCE_UNRESOLVED (cli#4766) catches both item 13 seeds: an
sla-status-change naming an undeclared SLA title, and the unclosed form from a
failed S4 run (`sla-status-change("Assess (`assess`)`). Tally: validate 22,
validate-warns 2, convert 6, excluded 1, silent 3 (17-19), needs-draft 4
(30-33). cli c2b3c72db (the #4743 rollup tip, adds #4774) changes no gate check
or convert output, so the verdicts hold there.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…th arms

runnability_disclosed builds the all-placeholder sla_from_sdd case and grades
that the completion report says the case cannot run and names where it stops.
runnability_silent builds the runnable expense case (all resources deployed)
and grades that the report says nothing about runnability. Each arm's
precondition is deterministic (check_runnability_shape.py --expect
blocked|runnable, falsified on a 29-placeholder user case and a resolved
fixture); only the reply is judged.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… line, not a numbered report field

runnability_silent failed 2/2: on a fully runnable case both completion
reports printed a Runnability section with zero counts, although field 6 said
to omit it at zero. A numbered field reads as a slot to fill. The disclosure now
sits after the field list as two lines written only when a count is non-zero,
with an explicit 'write nothing' for the clean case.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…ity items

cli 6f3b4ba96 (the #4743 rollup; items 17-19 in cli#4766 at 7619dd18b) puts
data closure behind sdd validate: SDD_VARIABLE_NEVER_PRODUCED,
SDD_OUT_VARIABLE_UNSET, SDD_BUTTON_TARGET_UNKNOWN. Coverage over all 35
items: validate 25, validate-warns 2, convert 6, excluded 1, needs-draft 4,
silent 0. The planner now checks only items 30-33, and only when finalizing a
draft; checklist item 17 also names a Buttons assignment as a producer, as
the gate does.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…; never stamp a marker into one resolve wrote

#3671's "never change a field resolve wrote" did not move the refusal rate
(3/6 runs before, 3/6 after, hitl smoke, cli f7c68a2e9): planning §3.4 still
told the agent to record <UNRESOLVED: ...> in the entry's taskTypeId slot, and
Step 9.1 / placeholder-tasks / Rule 9 keyed the placeholder on that marker. So
the agent opened resolve's entry to stamp it, re-ran the lookup under the task
name ("37 entries checked"), rewrote searchQuery.name and cacheFile, and
convert --resolved refused. Resolve already writes selected:null and a
rationale for an empty lookup; that is now the placeholder signal, and the
marker belongs only to entries the agent writes itself.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…verbatim, not by reference

runnability_disclosed passed 2 of 4: the misses listed every placeholder but
never said a run would stop or where. The report paragraph now carries the two
lines as a fill-in template instead of pointing at the Phase 6 section.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…le, receipt or not; normalize only what the CLI cannot read

#3690 sent every SDD without the planner receipt to planner normalization.
Most case-eval SDDs (44 of 49) are real designs written outside the planner,
and the route fired inconsistently on them (init-validate normalized,
case_exit_full and timer_connector_from_sdd built directly). Rule 2 now
decides three ways: receipt present → build; no receipt but sdd parse reads
it whole → build it directly, exactly the same; no receipt and the parse
fails → normalize through the planner, keeping sdd.original.md.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…interrupts; parallel oversight is start-task or notify-only

The docs taught a non-interrupting secondary SLA "oversight lane"
(Interrupting: No). The case compiler sets cancelStagesAndTasks for any
secondary-stage entry, whatever isInterrupting says, and cli#4804 now refuses
the shape. Falsified 2026-10-05 rather than taken on report: an
isInterrupting:false secondary case-entered entry packs to
"cancelStagesAndTasks":true, and debug instance 6ecbe140 ran only the
secondary lane while the primary stage, also on case-entered, never started.

Rule 22, the secondary-lane anti-pattern, sla-response-shapes §1/§4/§5,
the stages and stage-entry plugins, the planner layers guide (carve-out,
response table, Interrupting cell) and the template's secondary rule now say
Yes on every secondary entry, and route work-keeps-running requirements to
start-task or notify-only.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… an at-risk SLA rule

All 14 sla-status-change rows in the eval fixtures were 2-argument breach rows,
so nothing exercised the template's 3-argument at-risk form. This variant of
sla_from_sdd moves Intake's start-task to at-risk and grades the built rule:
Intake's slaId plus the escalationId of the 'Intake at risk' at-risk
escalation. Falsified on cli 6f3b4ba96, which read argument 3 as the SLA
title: convert left the rule with no slaId at all.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… head (cli 90b5247e6)

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…d-stage-exited carries an IF

A secondary stage entered on selected-stage-exited(X) with no IF takes over
every time X completes normally. Seen in a planner run (loan-origination,
Withdrawn lane on any exit of Intake and Loan Setup) and in 10 corpus
documents. cli#4766 now refuses it as SDD_SECONDARY_ENTERS_ON_COMPLETION
(item 36). The checklist (the planner's no-sdd fallback walk), the template
count and the coverage seed follow; coverage on cli fe4c3e3e2 reads it validate.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…ceipt-carrying SDD; without the receipt it normalizes

Rule 2's three-way decision sends an SDD with no planner receipt that fails
sdd parse to planner normalization (c), but its refusal clause said any
parse/format/resolve/convert refusal stops the build. The suite showed which
one wins: all four no-receipt SDDs refused on a Case Variables Type `number`
(io-binding-matrix, three hitl-quality tasks) stopped and asked the user,
scoring 0.00 at $0.18-0.46. The stop now names case (a) only, and (c) names a
single-cell refusal as its own. E1a's fixture carries the receipt, so its
hand-back is unchanged.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…lookup fields are untouchable

registry-handoff-ambiguous passed 14/14 nightlies on main and failed both
stack runs (0.14, 0.57). #3671's "never change a field resolve wrote" listed
`selected`, so after the user picked a candidate the agent recorded a
gateDecision but left selected null and bound nothing. convert --resolved
compares stage, task and searchQuery, not selected: the rule now protects the
lookup fields and names the one write a gate answer requires (selected to the
chosen match, with gateDecision pick:<name>). An ambiguous entry goes to the
gate before it can become a placeholder.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…ompleted

The Lifecycle gates table allowed only required-tasks-completed and
wait-for-connector on a completing stage row, matching the CLI gate's item 9.
Both were wrong (cli-62, from @uipath/case-schema: the schema, compiler and
scheduler all accept selected-tasks-completed as a TasksCompleted completion
condition; cli fixed in 990b37cda). Planner normalization rewrote a real SDD's
selected-tasks-completed("Approve Contract") completion row to
required-tasks-completed, changing which tasks complete the stage, and
hitl quality_04 failed on it. Checklist item 9 reads this table, so the
fallback walk follows.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…gainst runtime truth

case_breach_oversight graded isInterrupting false, which the runtime falsified
(debug instance 72162d9b: the breach exited the active stage and cancelled its
task). It now requires isInterrupting true on the still-secondary, non-required
lane and judges whether the reply tells the user the work will pause.

quality_01 accepts one trivial required task in Closed: an empty stage with a
required-tasks-completed exit never completes (instance cc7223dc).

io-binding accepts double or integer where the SDD wrote number, the form
planner normalization produces.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…et layout

The SDD declares no task set, so convert puts each timer, with its own
current-stage-entered rule, in its own set. Debug instance 2df59883 shows two
timers in separate sets start and finish together, so the shared-set
assertion graded layout, not behaviour. The entry-rule assertion stays.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The oversight eval accepted only a lane, but the skill's parallel-oversight
shape is a start-task. Debug instance 583c4102 shows a root-SLA breach during
Intake starts such a task even in not-yet-entered Decision, cancels nothing, and
the case completes, so it truly runs alongside. The checker now passes either a
non-required breach task with no new stage, or a secondary lane with
isInterrupting true; the judge grades the reply's claim against the shape built.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…ng input

The io_binding fixture writes the non-template type number on purpose: its eval
grades the no-receipt, parse-fails route to planner normalization (run
37494190705: the planner rewrote exactly those three cells, build 1.00). None of
the parse-clean test's three verdicts fit, and an exclusion would stop checking
the file. A fourth verdict, routing-input, pins the exact refusal instead, so a
new defect in the file or a CLI that starts accepting number both fail.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…st their seed

read-as-seed required sdd validate to exit zero, but the seeds are hand-written
pre-template SDDs that strict validate rejects on their own (32 issues on
e2e_expense_runnable: no Planner Handoff, task envelopes or activation modes).
The fixture now must give the seed's exit code and issues plus exactly its
recorded warning, and the seed's Model. table-row-missing-leading-pipe passes
on 1.205: the row and everything below it are read, with the warning at 312.
Records the 1.205 measurement beside the old reader's.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…defect

The norm() merge it was built for is fixed; what remained was a non-identifier
variable name that convert wrote as an unreachable id, retyped to string, and
both validators passed. cli #4804 now refuses it at line 70 with
SDD_VARIABLE_NAME_NOT_IDENTIFIER, so reject-naming-line passes as recorded.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…names cells

A no-receipt SDD one cell from parsing (a Case Variables Type of number) took
the full re-render: about 150 s on Opus, 20 s of it re-typing the original and
75 s regenerating a document that needed one word changed. Nine of the ten
normalizing evals were cell-level. The lane now copies the original with one
cp, edits only the named cells when parse read the stages, and hands back once
it parses whole (case Rule 2 (b)); the template render runs only for a refusal
that is not cell-level.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…from sdd format

The test hand-built a sidecar with no schemaVersion; the CLI's model schema
moved to 2.0.0 and refuses an unversioned file as 1.0.0. The envelope now
comes from a real sdd format run, so the test follows the CLI's schema.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
On Opus the design lane's gate got 40% faster once sdd validate replaced the
checklist walk, but the writing phase gave it back: both finalize tasks ran
validate twice on half-written files and repaired errors the next append would
have cleared, and 4 of 6 design tasks ran --help or --version to rediscover
commands the lane already states.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@roberts-cliff

Copy link
Copy Markdown
Contributor Author

Closing the MST-15966 SDD-quest stack; it isn't landing as is. The head is kept locally.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant