codex-workflow-v2 2.0.0-beta.12.10 → 2.0.0-beta.12.12

package/docs/validation-report.md

@@ -1,105 +1,134 @@
-# V2 beta.12.10 Validation Report
+# V2 beta.12.12 Validation Report
 
-Release candidate `2.0.0-beta.12.10` was validated on 2026-08-31 (Europe/Moscow) with
+Release candidate `2.0.0-beta.12.12` was validated on 2026-08-31 (Europe/Moscow) with
 Node.js 24.17.0.
 
 ## Automated release evidence
 
-- `npm ci` completed from the exact lockfile with 0 reported vulnerabilities.
-- `npm run validate` passed typecheck and `344/344` sequential tests with no failures, skips,
-  cancellations, or todos in 2,087,253.659208 ms.
-- `npm run test:signal-remediation-e2e` passed the focused Signal-shaped end-to-end scenario.
-  The test starts from an active Milestone whose required Task has a current `stop-escalate` Plan
-  Risk Audit, materializes one audited upstream remediation Task, completes its Plan/audit/C1/Step/
-  review/accept/merge lifecycle, replans the original Task, and completes a later Step there.
-- Negative coverage rejects a non-exact predecessor frontier, duplicate remediation, stale revisions,
-  non-required or non-active Milestone posture, topology drift, reused audit evidence, and ordinary
-  standalone/post-execution `stop-escalate` cases that must keep the existing `task plan-set` route.
-- The complete lifecycle inventory names `milestone remediation-materialize` and
-  `WorkflowService.materializeAuditedRemediationTask`; read-only purity, transaction rollback/
-  roll-forward, concurrent writers, credentials, Core receipts and prior Signal compatibility tests
-  all passed.
-- Package dry-run produced `codex-workflow-v2-2.0.0-beta.12.10.tgz` with 278 files, 1,050,219 bytes
-  packed and 4,196,113 bytes unpacked. The broad `docs` package allowlist was narrowed so the
-  repository-local `docs/quality` observation journal cannot enter the tarball. The first invocation
-  encountered a root-owned user npm cache (`EPERM`); the exact retry used an isolated temporary cache.
-- `npm run smoke:downstream` built and installed the tarball in a disposable consumer project and
+- `npm ci` completed from the exact lockfile; the audit reported zero vulnerabilities.
+- The first full `npm run validate` run found two compatibility regressions in the pre-existing
+  corrective-recovery suite. Plan-integrity preflight was made non-authoritative while a repairable
+  legacy remediation-history defect exists, and an old test was aligned with the intentional
+  corrective-yield-before-context-refresh priority. Both focused reproductions then passed.
+- The mandatory second `npm run validate` passed typecheck and `348/348` sequential tests with no
+  failures, skips, cancellations, or todos in 2,149,397.21 ms.
+- The full suite includes the beta.12.12 selector-contract E2E: four sequential failures of one
+  guarded Step, dependency-only update transport, exact compatibility registration, automatic
+  package-only Knowledge refresh, Plan-integrity recovery, corrective yield, and fresh corrective
+  replan without losing the Step worktree.
+- `npm run test:signal-remediation-e2e` passed both focused scenarios (`2/2`). It still proves the
+  audited upstream-remediation route through a later Step and rejects unrelated Knowledge drift.
+- Package dry-run produced `codex-workflow-v2-2.0.0-beta.12.12.tgz` with 281 files, 1,070,498 bytes
+  packed and 4,283,430 bytes unpacked. The tarball SHA-1 is
+  `349bcd371b09a2467496057166c78d879e99e937`. The first invocation encountered a root-owned user npm
+  cache (`EPERM`); the exact retry passed with an isolated temporary cache and no system permission
+  changes.
+- `npm run smoke:downstream` built and installed that tarball in a disposable consumer project and
   returned `ok=true`. The sandboxed attempt could not resolve npm (`ENOTFOUND`); the same command
-  passed with network access.
-- `npm run plugin:check` passed for bundled gateway
-  `0.1.0+codex.20260830183000`.
-- `npm run docs:pdf:check` reported an empty stale set and exact beta.12.10 package binding.
-- `RELEASE_TAG=v2.0.0-beta.12.10 npm run release:check` and `git diff --check` passed after this
-  report and the exact package allowlist were recorded.
-
-## Root cause corrected by beta.12.10
-
-An active Milestone could reach a genuine pre-execution dead end when a required Task's current Plan
-Risk Audit truthfully returned `stop-escalate` because an upstream capability did not exist. The old
-choices were insufficient: replanning the same Task could not manufacture the missing capability,
-while the generic scope-change path required a separate human approval and permitted a much wider
-semantic change than the evidence justified.
-
-The problem was missing bounded topology authority, not an insufficient number of implementation or
-review attempts. Repeating the blocked Task could only reproduce the same unmet prerequisite.
+  passed with approved network access.
+- `npm run plugin:check` passed for bundled gateway `0.1.0+codex.20260830183000`. The installed
+  Codex CLI exposes no `plugin validate` subcommand, so the repository validator plus successful
+  packaged downstream installation are the available plugin release evidence.
+- `npm run docs:pdf:check` reported an empty stale set and exact beta.12.12 package binding.
+- `RELEASE_TAG=v2.0.0-beta.12.12 npm run release:check` and `git diff --check` passed.
+- `node scripts/validate-lifecycle-transcript.mjs --file
+  tests/fixtures/lifecycle-transcripts/valid-normal-sequence.jsonl` passed with four checked events;
+  the full validator suite also passed all positive and negative transcript fixtures.
+
+## Root cause corrected
+
+Signal Task `TASK-01M1B79C660YMNSRRZ236NTR1T` reached revision 12 with four recorded
+`checks-failed` events. The approved Step check was `npm run validate:affected`, while the existing
+repository script delegates to `tools/validation/run.mjs affected`, whose public contract requires at
+least one of `--base`, `--head`, or `--files`. The command therefore failed before validation and could
+not become successful through another implementation attempt.
+
+Two Workflow gaps made the state non-recoverable:
+
+1. mechanical feasibility proved that the npm script existed but did not verify the mandatory
+   selector contract of the repository-owned runner; and
+2. bounded Plan-integrity recovery recognized only the first failed attempt with no corrective
+   decision, so it could not convert the already accumulated four-event history into a fresh Plan.
+
+The primary defect was an invalid executable Plan contract plus an incomplete late-recovery route,
+not insufficient Worker effort and not an attempt-count problem.
 
 ## Implemented lifecycle contract
 
-- Fresh `next` advertises audited remediation only for an `awaiting_execution_authorization`
-  required Task of an active Milestone with a current `stop-escalate` Plan Risk Audit. Standalone,
-  post-execution, `split-required`, and other stop postures preserve their existing routes.
-- Coordinator first records a normal capability Discovery. `milestone remediation-materialize`
-  then performs one project transaction that appends exactly one required remediation Task and adds
-  exactly one dependency edge from the blocked Task to it.
-- The new Task must inherit the blocked Task's exact direct predecessor frontier, and every predecessor
-  must already be merged. Milestone outcome, success signal, acceptance, checks, base, old memberships
-  and all unrelated dependency edges remain byte-semantically unchanged.
-- The scope-change event binds Milestone revision, blocked Task revision, Brief/Plan hashes, exact Plan
-  Risk Audit event/hash, remediation Discovery revision, new Task identity and predecessor frontier.
-  Reuse or topology drift fails closed.
-- A `corrective-derived` authorization, bound to the audited evidence and exact actor
-  `agent:milestone-remediation:<MS-ID>`, replaces the superseded Milestone execution authorization.
-  It authorizes only current required Task execution acceptance transitions; it is neither a general
-  delegation grant nor human impersonation.
-- The remediation route requires no human approval. After its Task merges, the original Task returns
-  to `task plan-set`, must receive a semantically new Plan and fresh Plan Risk Audit, and then proceeds
-  through the ordinary C1, checks, commits and reviews.
-- Public handshake capabilities include `audited-remediation-topology-v1` and
-  `corrective-derived-task-authority-v1`.
-
-## Signal-shaped evidence
-
-The release E2E mirrors the observed Signal deadlock: four already merged provider Tasks, one blocked
-required consumer Task, a current pre-execution `stop-escalate` audit, and a missing upstream
-capability. It proves the entire recovery cycle rather than only the new materialization command.
-
-Live Signal installation and mutation are intentionally separate post-publication actions. They must
-use exact repository-local beta.12.10, a clean/idle update boundary, fresh `status -> next`, and the
-advertised compatibility/recovery route if the dependency update advances active Milestone base
-state. No manual state edit, Git history rewrite, Task cancellation, or duplicate Milestone is part of
-the release mechanism.
+- Mechanical feasibility now uses the versioned analyzer `wf2.root-npm-script-arguments@1`. It
+  conservatively recognizes only a root npm script whose command is a repository-relative Node
+  `affected` runner and whose source contains the established selector contract. A Plan check without
+  `--base`, `--head`, or `--files` is blocked when the runner is outside current/predecessor
+  `allowedWrites`; an explicit forwarded selector passes. Unsupported syntax remains `unverified`.
+- Plan-integrity evidence schema v2 binds the current Task revision, Step and Plan hashes, every
+  sequential remediation event, package manifest and runner bytes, Git HEAD, dirty Step files, and a
+  canonical conflict hash. Existing v1 evidence remains readable.
+- Recovery accepts one or more contiguous `checks-failed` events only when no corrective decision has
+  consumed them. It appends one `replan-required` decision at the next ordinal and covers the complete
+  failed history; it does not synthesize an extra failure.
+- A published update from beta.12.11 is permitted only through the read-only
+  `update plan-integrity-update-preflight` proof. Transport may commit only `package.json` and
+  `package-lock.json` on the active Milestone base and carry that exact dependency commit to the Task
+  branch while preserving all product bytes.
+- Fresh `next` then advertises `update plan-integrity-dependency-recover` only when current HEAD,
+  package version, parent commit, Task/Step/Plan, remediation IDs, dirty paths and their content hash,
+  absence of leases, and transaction state still match the preflight boundary.
+- Dependency registration and a Project Knowledge refresh are one Project transaction. The refresh is
+  allowed without a Human gate only when the sole canonical drift is the exact Workflow pin in
+  `package.json`; any unrelated Knowledge change fails closed.
+- The compatibility record is authority only for the exact post-update Plan-integrity recovery. A
+  fresh replacement Plan must bind the current Knowledge Map and replace the invalid check with an
+  executable selector-bearing command. Ordinary authorization, C1, Core commit, strict review, and
+  subsequent navigation remain mandatory.
+- A repairable legacy remediation-mode defect outranks this new preflight. Plan-integrity assessment
+  reports itself ineligible and leaves `next` free to advertise the established
+  `task remediation-mode-recover` route.
+- Handshake adds `plan-integrity-update-recovery-v1`; protocol and state schema remain `2/2`.
+
+## Signal-shaped and negative evidence
+
+The new synthetic E2E reproduces the real Signal boundary rather than testing a helper in isolation:
+
+- beta.12.11 package/lock and approved Knowledge Map;
+- an active claimed guarded Step with four distinct `checks-failed` events;
+- the same missing selector contract as Signal;
+- the exact allowlisted dirty worktree;
+- a dependency-only beta.12.12 commit;
+- automatic package-only Knowledge refresh and compatibility registration;
+- writer credential replacement, Plan-integrity recovery, corrective yield, and fresh corrective
+  replan preparation.
+
+Negative coverage rejects a runner inside the current Step write scope, unrelated dirty files,
+tampered evidence, an active lease, wider dependency commits, changed worktree bytes, stale
+Task/Plan/Step bindings, transaction conflicts, and unrelated Knowledge drift. Existing tests also
+prove that Plan-integrity preflight cannot hide the older remediation-mode recovery route.
+
+Live Signal mutation remains post-publication work. It must preserve the three current Worker files,
+transport only package/lock, use fresh repository-local handshake/status/next before each mutation,
+and follow only advertised recovery actions. No manual Workflow-state edit, history rewrite, Task
+cancellation, duplicate Task, or replacement Milestone is authorized by this release.
 
 ## PDF release documentation evidence
 
-- All three tracked Russian PDFs were regenerated from beta.12.10 Markdown sources.
-- Architecture is 8 pages, the delegated chat-only guide is 10 pages, and the technical reference is
-  12 pages. All 30 pages were rendered and visually inspected; the changed recovery sections were
-  checked at full-page resolution.
-- Titles, package bindings, Cyrillic text, tables, code blocks, margins, footers and section transitions
-  have no visible clipping, overlap or unreadable content.
+- All three tracked Russian PDFs were regenerated from beta.12.12 Markdown sources.
+- Architecture is 9 pages, the delegated chat-only guide is 10 pages, and the technical reference is
+  13 pages. All 32 pages were rendered and visually inspected through contact sheets.
+- Titles, version bindings, Cyrillic text, tables, code blocks, margins, footers and selector-recovery
+  sections show no clipping, overlap, or unreadable content.
 
 ## Existing boundaries and non-claims
 
-- The new route does not permit arbitrary Task insertion, task removal, semantic Milestone scope
-  change, arbitrary dependency rewiring, completed-Step adoption or `split-required` replacement.
-- Human approval is removed only from this exact evidence-derived corrective topology transition.
-  Final Milestone acceptance and generic semantic scope changes retain their existing authority model.
-- Protocol version remains 2 and state schema remains 2; older projects remain readable without a
-  state migration.
-- No Git commit, tag, remote push, npm publication or live Signal recovery is claimed by this
-  pre-publication report.
-
-Confidence is high for the package implementation, topology invariants, no-human corrective authority,
-regression compatibility, tarball composition, downstream installation, plugin and documentation.
-Live Signal confidence remains conditional on successful publication, exact dependency installation,
-fresh Core readback and Coordinator completion of the advertised remediation route.
+- The analyzer does not execute or infer arbitrary shell grammar and does not claim to understand
+  every repository validation framework.
+- The compatibility route cannot change product files, Plan semantics, Milestone topology, Task
+  membership, dependency edges, completed-Step provenance, or unrelated Knowledge.
+- The release does not make a failing implementation pass, approve a broken check, or increase retry
+  limits. It creates a bounded route to replace a mechanically impossible approved check.
+- Protocol version remains 2 and state schema remains 2; no state migration is required.
+- This report does not claim Git commit/tag/push, npm publication, live Signal installation, or Task
+  completion before those operations receive their own readbacks.
+
+Confidence is high for selector detection, repeated-failure recovery, update compatibility,
+transaction atomicity, regression compatibility, package composition, downstream installation,
+gateway and documentation. Live Signal confidence remains conditional on exact publication,
+installation, fresh Core readback, and completion of the previously blocked Step.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "codex-workflow-v2",
-  "version": "2.0.0-beta.12.10",
+  "version": "2.0.0-beta.12.12",
   "description": "Requirements-first Codex development workflow with local state and explicit Git ownership.",
   "repository": {
     "type": "git",

package/plugins/codex-workflow-gateway/skills/codex-workflow-gateway/SKILL.md

@@ -156,6 +156,16 @@ free-form Milestone scope change. Follow only the advertised route:
    record a fresh Risk Audit, use the same current derived approval only when advertised, and
    continue through at least the next ordinary Step boundary.
 
+An exact package update may make the approved Knowledge Map stale while this route is active. Continue
+without a Human approval only when fresh `next` includes
+`auditedRemediation.dependencyKnowledgeRefresh.eligible=true`. Core must prove that `package.json` is
+the sole changed Knowledge source, every version surface equals the repository-local runtime, and every
+Git commit that changed the approved manifest content altered only the exact Workflow dependency in
+`package.json`/`package-lock.json`. Do not run standalone reconcile first. The refresh is committed
+atomically by `milestone remediation-materialize` with corrective-derived evidence. Any other Knowledge
+path, classification, gap, conflict, dirty checkout, unaligned version or wider commit returns to the
+ordinary explicit Knowledge flow.
+
 This is an idempotent, additive topology transition bound to one exact Plan Risk Audit. Dirty Git,
 any lease/transaction/Core operation, a started blocked Task, stale Task/Milestone/audit bindings,
 non-merged or non-exact predecessors, duplicate use of the audit, or any changed Milestone semantic

package/schemas/corrective-decision-event.schema.json

@@ -50,11 +50,16 @@
       "then": {
         "required": ["recoveryBasis", "planIntegrityEvidence"],
         "properties": { "decision": { "const": "replan-required" } }
-      },
-      "else": {
-        "not": { "anyOf": [{ "required": ["recoveryBasis"] }, { "required": ["planIntegrityEvidence"] }] }
       }
     },
+    {
+      "if": { "required": ["recoveryBasis"] },
+      "then": {
+        "required": ["planIntegrityEvidence"],
+        "properties": { "decision": { "const": "replan-required" } }
+      },
+      "else": { "not": { "required": ["planIntegrityEvidence"] } }
+    },
     {
       "anyOf": [
     { "properties": { "coveredFindingIds": { "minItems": 1 } } },
@@ -74,13 +79,14 @@
       "items": { "type": "string", "minLength": 1 }
     },
     "planIntegrityEvidence": {
-      "type": "object",
-      "additionalProperties": false,
-      "required": [
-        "kind", "taskRevision", "remediationEventId", "remediationEventHash",
-        "manifestPath", "manifestHash", "gitHead", "stepDefinitionHash",
-        "missingScripts", "checkCommands", "changedFiles", "conflictHash"
-      ],
+      "oneOf": [
+        { "$ref": "#/$defs/planIntegrityEvidenceV1" },
+        { "$ref": "#/$defs/planIntegrityEvidenceV2" }
+      ]
+    },
+    "planIntegrityEvidenceV1": {
+      "type": "object", "additionalProperties": false,
+      "required": ["kind", "taskRevision", "remediationEventId", "remediationEventHash", "manifestPath", "manifestHash", "gitHead", "stepDefinitionHash", "missingScripts", "checkCommands", "changedFiles", "conflictHash"],
       "properties": {
         "kind": { "const": "missing-npm-script-outside-allowed-writes" },
         "taskRevision": { "type": "integer", "minimum": 1 },
@@ -95,6 +101,47 @@
         "changedFiles": { "$ref": "#/$defs/stringList" },
         "conflictHash": { "$ref": "#/$defs/hash64" }
       }
+    },
+    "planIntegrityEvidenceV2": {
+      "type": "object", "additionalProperties": false,
+      "required": ["kind", "taskRevision", "remediationEvents", "manifestPath", "manifestHash", "gitHead", "stepDefinitionHash", "conflicts", "changedFiles", "conflictHash"],
+      "properties": {
+        "kind": { "const": "non-executable-npm-check-outside-allowed-writes" },
+        "taskRevision": { "type": "integer", "minimum": 1 },
+        "remediationEvents": {
+          "type": "array", "minItems": 1,
+          "items": {
+            "type": "object", "additionalProperties": false,
+            "required": ["eventId", "eventHash", "attemptOrdinal"],
+            "properties": {
+              "eventId": { "type": "string", "minLength": 1 },
+              "eventHash": { "$ref": "#/$defs/hash64" },
+              "attemptOrdinal": { "type": "integer", "minimum": 1 }
+            }
+          }
+        },
+        "manifestPath": { "const": "package.json" },
+        "manifestHash": { "$ref": "#/$defs/nullableHash64" },
+        "gitHead": { "$ref": "#/$defs/gitOid" },
+        "stepDefinitionHash": { "$ref": "#/$defs/hash64" },
+        "conflicts": {
+          "type": "array", "minItems": 1,
+          "items": {
+            "type": "object", "additionalProperties": false,
+            "required": ["reasonCode", "script", "command", "scriptCommand", "runnerPath", "runnerHash"],
+            "properties": {
+              "reasonCode": { "enum": ["MISSING_NPM_SCRIPT", "REQUIRED_SELECTOR_MISSING"] },
+              "script": { "type": "string", "minLength": 1 },
+              "command": { "type": "string", "minLength": 1 },
+              "scriptCommand": { "type": ["string", "null"] },
+              "runnerPath": { "type": ["string", "null"] },
+              "runnerHash": { "$ref": "#/$defs/nullableHash64" }
+            }
+          }
+        },
+        "changedFiles": { "$ref": "#/$defs/stringList" },
+        "conflictHash": { "$ref": "#/$defs/hash64" }
+      }
     }
   }
 }

package/schemas/task.schema.json

@@ -177,6 +177,29 @@
               "retainedCommitSha": { "type": "string", "pattern": "^[a-f0-9]{40}$" }
             },
             "additionalProperties": false
+          },
+          "planIntegrity": {
+            "type": "object",
+            "required": ["stepId", "planHash", "conflictHash", "dirtyFiles", "dirtyWorktreeHash", "remediationEventIds"],
+            "properties": {
+              "stepId": { "type": "string", "pattern": "^STEP-[A-Z0-9][A-Z0-9-]{2,63}$" },
+              "planHash": { "type": "string", "pattern": "^[a-f0-9]{64}$" },
+              "conflictHash": { "type": "string", "pattern": "^[a-f0-9]{64}$" },
+              "dirtyFiles": {
+                "type": "array",
+                "minItems": 1,
+                "uniqueItems": true,
+                "items": { "type": "string", "minLength": 1 }
+              },
+              "dirtyWorktreeHash": { "type": "string", "pattern": "^[a-f0-9]{64}$" },
+              "remediationEventIds": {
+                "type": "array",
+                "minItems": 1,
+                "uniqueItems": true,
+                "items": { "type": "string", "minLength": 1 }
+              }
+            },
+            "additionalProperties": false
           }
         },
         "additionalProperties": false