@nebulacomponents/citable 1.11.0 → 1.13.0

package/CHANGELOG.md

@@ -2,6 +2,157 @@
 
 ## Unreleased
 
+## 1.13.0 — 2026-07-19
+
+### Added — Release Evidence Governance
+
+- Canonical release manifests that bind executable detector, namespace,
+  registry, provider, distribution, documentation, and generated publisher
+  projections to the release commit with fail-closed checksum validation.
+- Two-phase release governance with `published_unfinalized`, immutable
+  finalization, bounded dwell, terminal withdrawal or supersession, and
+  owner-controlled deployment receipts that cannot be mistaken for independent
+  attestation.
+- Direct and cache-busted representation-drift observations with longitudinal
+  divergence and convergence evidence that remains `external_unverified` and
+  cannot satisfy release gates.
+- Acceptance-corpus contracts and bounded accuracy metrics covering detector
+  confusion matrices, incomplete evidence, reviewer agreement, execution cost,
+  reproducibility, and remediation verification.
+- Acceptance-run reproducibility receipts that verify sealed run packages,
+  preserve partial and unavailable observations, exclude disclosed
+  environment-specific fields from fingerprints, and compare property,
+  detector, configuration, observation-method, tool, and external-system drift
+  independently.
+- Real-property corpus authorization, architecture, collection, retention,
+  sanitization, publication, and run-evidence contracts, plus a fail-closed
+  public projection with a hash-bound owner-authorized publication receipt.
+- Publishable field-validation metrics with corpus and collection provenance,
+  sample/census identity, detector-version cohorts, explicit numerator and
+  denominator definitions, unknown-false-negative state, confidence limits,
+  and hash-bound JSON and Markdown projections.
+- Schema-validated cross-browser and interaction-journey evidence plans with
+  explicit engine/version, device, JavaScript, locale, consent,
+  authentication, step, artifact, and partial-failure records.
+- Versioned, independently verifiable artifact interchange with fail-closed
+  `artifacts export`, `artifacts verify`, and `artifacts import` commands. The
+  contract preserves canonical run bytes and explicitly excludes hosted
+  tenancy, authentication, storage, billing, and dashboard authority.
+
+### Migration — Acceptance Corpus Schema v2
+
+- Acceptance corpus manifests must migrate `schema_version` from `1` to `2`.
+  Replace `owner_authorized` with the structured `authorization` record; add
+  `architecture`, `collection_methods`, `publication`, `sanitization`,
+  intentional-incompleteness and contradiction state; add run artifact and
+  governance references; and declare false-negative discovery and causal
+  limitations. Schema v1 manifests are rejected rather than inferred.
+  Corpus v2 manifests created before this change must also add
+  `evaluation_design`, and every detector case must add `detector_version`.
+
+### Changed — Release Evidence Governance
+
+- PR policy validation now skips post-merge edited events, preventing deleted
+  head references from creating misleading failures after a successful merge.
+- Controlled surfaces now declare exact-body or projection-hash-header
+  verification. Rendered pages without the declared header produce
+  `insufficient_evidence` instead of a false body-hash divergence.
+- Release shipping now creates a draft release, records npm publication as an
+  explicit unfinalized state, and requires valid controlled-surface receipts in
+  a separate protected workflow before publishing the GitHub release as latest.
+- Field-derived detector applicability now keeps non-HTML resources eligible
+  for transport checks while restricting HTML semantics to HTML documents,
+  excludes Cloudflare `/cdn-cgi/` utility routes from default index-target
+  intent, and reports bounded URL crawls as incomplete coverage instead of
+  treating unfetched sitemap URLs as confirmed errors.
+
+### Added — Executive Reporting Suite
+
+**Governed input registries (8 new schemas + registry specs)**
+
+- `kpi` command — KPI Architecture registry: govern metric definitions, sources,
+  targets, calculation, known limitations, and restatement policy. Refuses metrics
+  lacking executive definitions or empty `known_limitations`.
+- `variance` command — Variance analysis registry: enforces four-act narrative
+  (plan → actual → cause → response). Rejects vague causes (`market conditions`,
+  `timing`, `customer behaviour`) as primary drivers. Material/critical variances
+  require `management_response`.
+- `outcomes` command — Customer Outcomes registry: separates activity from
+  validated impact. Enforces `finding_produced → finding_accepted →
+  answer_engine_changed → commercial_value_claimed →
+  causal_relationship_established` progression. Blocks `commercial_value` at
+  `finding_produced` stage.
+- `risk` command — Risk Register: top risks sorted by residual exposure, KRI
+  tracking, trigger thresholds, control effectiveness. Board-visible risks require
+  `trigger_threshold`. High/critical risks require `key_risk_indicators`.
+
+**Executive output commands**
+
+- `executive-review` — Monthly Executive Operating Review. Evidence ledger built
+  first (governed metrics, variance counts, outcome stages, risk counts), then
+  four-act narrative template generated. Returns `ungoverned_warning` when no KPIs
+  registered — refuses to produce narrative from ungoverned data.
+- `board-report` — Quarterly Board Pack. Every reported statement carries
+  `statement_type`, `source_records`, `period`, `confidence`, and `limitations`.
+  Returns `refused_sections` when kpis/customer-outcomes/risks are absent.
+- `decision-memo` — Bounded Decision Record. Requires: named consulted parties,
+  `what_would_change_recommendation`, real trade-offs on every option (no
+  strawmen), specific `supporting_evidence`, and `reopen_conditions` for
+  effectively-one-way decisions.
+
+**Strategic analysis commands**
+
+- `assumption-audit` — Assumption validity tracking. Validated assumptions require
+  `evidence_for`; contradicted assumptions require `evidence_against`; critical
+  importance requires `next_test` date.
+- `scenario` — Compound Risk Scenario War Room. Capped at 3 variables per
+  scenario. Requires `cascade_analysis`, all three states (base/stress/severe),
+  hedges with cost/impact/owner/deadline, and `early_warning_triggers`.
+- `prioritize` — Initiative Prioritization with transparent scoring: `(demand +
+  revenue + differentiation − engineering − operating − legal) × evidence_multiplier`.
+  Weights exposed in output. Rejects `revenue_potential=transformative` with
+  `evidence_strength=assumption`.
+- `competitive-intel` — Competitive Intelligence. Every claim requires
+  `observation_date`, `source`, and `claim_type`. Unreliable sources (G2, Capterra,
+  homepage, LLM recall) require `independent_verification`. Includes `stale`
+  subcommand.
+
+**Orchestration**
+
+- `executive` command — Chief-of-Staff Router. Routes 11 request patterns to
+  the correct reporting skill. Max orchestration depth: 1 (no recursive routing).
+  Logs all routing decisions to `.citable/executive-log.yaml` (last 100 retained).
+  `--route <command>` forces a specific route. `--log` shows recent decisions.
+
+**Infrastructure**
+
+- 8 new JSON schemas registered in `REGISTRY_SPECS`:
+  `kpi`, `variance`, `customer-outcome`, `risk`, `decision`, `assumption`,
+  `scenario`, `initiative` — all with `citable://schemas/` `$id` prefix for AJV
+  compatibility.
+- Comprehensive unit coverage in `tests/unit/executive-reporting.test.js` for
+  governed inputs, reports, cross-registry integrity, deterministic IDs, and fixed-date evaluation.
+
+## 1.12.0 — 2026-07-19
+
+### Fixed
+
+- Replaced regex-based HTML and transcript tag stripping with structured parsing
+  that excludes executable elements, including malformed end-tag variants.
+- Restricted schema-source discovery to exact `schema.org` hosts instead of
+  accepting deceptive hostname substrings.
+- Removed command-line-derived release-note regular expressions and an
+  ineffective identity replacement in schema integration coverage.
+
+### Added
+
+- Independent desktop, mobile, and JavaScript-disabled Chromium render profiles
+  with raw/rendered parity artifacts and bounded interaction discovery.
+- Partial-failure evidence and `--resume-run` recovery that reuses only
+  successful profiles from an immutable source run.
+- Optional repeated local Lighthouse execution with pinned runtime metadata,
+  per-run artifacts, lab/field separation, and median metric summaries.
+
 ## 1.11.0 — 2026-07-19
 
 ### Added

package/README.md

@@ -155,6 +155,9 @@ citable observe logs --input edge-logs.json
 citable observe bing --dataset ai_performance --input bing-ai-performance.csv
 citable observe consensus --target ./dist --base-url https://example.com
 citable observe performance --target https://example.com # CRUX_API_KEY
+citable observe performance --target https://example.com --lighthouse --repeat 3
+citable observe render --target https://example.com --interactions
+citable observe render --target https://example.com --resume-run <run-id>
 citable apply --input remediation-spec.json          # dry run
 citable apply --input remediation-spec.json --write  # reviewed + hash-locked
 citable monitor [runA runB]
@@ -174,6 +177,14 @@ citable reviews evaluate
 citable schedules run <schedule-id> --ref-date 2026-07-19
 citable project github <run-id>
 citable observe media --input media-manifest.json # add --ocr explicitly when required
+citable observe representation --input release-manifest.json --target https://example.com/llms.txt
+citable corpus evaluate --input acceptance-corpus.json
+citable corpus publish --input acceptance-corpus.json --output public-corpus.json
+citable corpus receipt --run <run-id> --input execution-context.json
+citable corpus compare-receipts <receipt-a.json> <receipt-b.json>
+citable artifacts export <run-id> --output ./portable-run
+citable artifacts verify --input ./portable-run
+citable artifacts import --input ./portable-run
 citable substantiate          # claim/evidence assessment (dry run)
 citable validate              # registry schema + referential integrity
 ```
@@ -184,7 +195,7 @@ citable validate              # registry schema + referential integrity
 | --- | --- |
 | `skill/` | Canonical agent skill: SKILL.md, command contracts, rubrics, anti-patterns, policies, templates |
 | `src/` | CLI, commands, registries, detectors, crawler/extractor, evidence, reporting |
-| `schemas/` | JSON Schemas: 19 registries plus findings, runs, observations, remediation, config, and prompt results |
+| `schemas/` | JSON Schemas: 27 registries plus findings, runs, observations, release governance, remediation, config, and prompt results |
 | `tests/` | Unit + integration suites; positive/negative fixtures |
 | `docs/` | ADR, traceability matrix, known limitations |
 | `dist/` | Generated distribution packages (run `npm run build:dist`) |
@@ -208,11 +219,27 @@ citable validate              # registry schema + referential integrity
 - [Known limitations](docs/known-limitations.md) includes medium-specific PDF,
   transcript, image-context, and optional OCR boundaries.
 
+## Community And Governance
+
+- [Contributing](CONTRIBUTING.md) - engineering and pull-request contracts
+- [Support](SUPPORT.md) - Issues, Discussions, hosted-service, and security routing
+- [Governance](GOVERNANCE.md) - authority, decisions, roles, and continuity
+- [Maintainers](MAINTAINERS.md) - current responsibilities and succession
+- [Security](SECURITY.md) - private reporting and supported-version policy
+- [Code of Conduct](CODE_OF_CONDUCT.md) - participation and enforcement
+
+Repository documentation and released schemas are canonical. Issues track
+accepted work, Discussions develop questions and proposals, and the Wiki holds
+non-authoritative operational recipes. The hosted Nebula Components dashboard
+is a separate product surface.
+
 ## Requirements
 
 Node >= 24.0.0. Network access is used for URL audits and explicitly selected
 live collectors. Render collection requires the optional `playwright` peer and
-an installed Chromium browser. Google index inspection uses `GSC_ACCESS_TOKEN`
+an installed Chromium browser. Local Lighthouse collection requires the
+optional `lighthouse` and `chrome-launcher` peers plus Chrome or Chromium.
+Google index inspection uses `GSC_ACCESS_TOKEN`
 and `--site-url`; CrUX uses `CRUX_API_KEY`. Missing dependencies or credentials
 produce incomplete evidence. Supported frameworks for auto-detection: Next.js,
 Astro, Nuxt, SvelteKit, Gatsby (other stacks: audit their built HTML output).

package/dist/universal/.agents/skills/citable/SKILL.md

@@ -7,7 +7,7 @@ description: >
   generative engine optimization, structured data governance, claim substantiation,
   crawler policy, entity consistency, content discoverability, or wants a site audited
   for how search and AI systems will retrieve, understand, cite, or recommend it.
-version: 1.11.0
+version: 1.13.0
 ---
 
 # Citable — evidence and change control for search and AI citation readiness
@@ -96,6 +96,9 @@ citable compare-snapshots [runA runB]
 citable action-plan [run-id]          # ordered actions, blockers, semantic gates, verification
 citable observe <mode> [options]      # render/index/citation/log/passage/consensus/performance evidence
 citable observe bing --dataset <search_performance|ai_performance> --input <csv|json>
+citable observe render --target <url> [--interactions] [--resume-run <run-id>]
+citable observe render --input <browser-evidence-plan.json>
+citable observe performance --target <url> --lighthouse [--repeat 1..5]
 citable apply --input <spec> [--write] # reviewed, hash-locked remediation; dry run by default
 citable monitor [runA runB]           # observation regression alerts
 citable metrics import --provider <name> --input <csv|json>
@@ -118,6 +121,14 @@ citable reviews evaluate
 citable schedules run <schedule-id> [--ref-date YYYY-MM-DD]
 citable project github <run-id>
 citable observe media --input <manifest.json> [--ocr]
+citable observe representation --input <release-manifest.json> --target <controlled-url>
+citable corpus evaluate --input <acceptance-corpus.json>
+citable corpus publish --input <acceptance-corpus.json> --output <public-corpus.json>
+citable corpus receipt --run <run-id> [--input <execution-context.json>]
+citable corpus compare-receipts <receipt-a.json> <receipt-b.json>
+citable artifacts export <run-id> --output <directory>
+citable artifacts verify --input <directory>
+citable artifacts import --input <directory>
 ```
 
 Audit scopes: `technical seo aeo geo architecture entity claims evidence schema
@@ -162,6 +173,8 @@ operator data or future adapters.
 Use `commands/observe-and-act.md` for collector prerequisites, evidence labels,
 the controlled citation adapter protocol, remediation refusal conditions, and
 monitoring interpretation.
+Use `commands/field-validation.md` for acceptance-corpus authorization,
+sanitization, publication refusal, reproducibility, and evidence boundaries.
 
 ## Command workflows
 

package/dist/universal/.agents/skills/citable/VERSION

@@ -1 +1 @@
-1.11.0
+1.13.0

package/dist/universal/.agents/skills/citable/commands/README.md

@@ -34,8 +34,9 @@ intent, not executability, so statuses are strict:
 | /citable ingest, map-site, map-queries, map-prompts, map-entities, map-evidence | specified (ingest.md) |
 | /citable optimize-page | specified (optimize-page.md) — requires source-to-render mapping, claim-preserving rewrites, build execution, rollback; none of that is proven yet |
 | /citable create-page, answer-block, architect, interlink, consolidate, metadata | specified (page-work.md) |
-| /citable validate-render | partially implemented by `observe render`; mobile/cross-browser and interaction exploration remain |
+| /citable validate-render | partially implemented by `observe render`; fixed Chromium profiles and schema-validated Chromium/Firefox/WebKit journeys are implemented, while reviewed semantic-impact detectors and reusable application-specific journey libraries remain |
 | /citable measure seo/aeo/geo, test-prompts, monitor-crawlers, monitor-contradictions | partially implemented by metric/observation imports, objective evaluation, Google/CrUX live APIs, custom citation adapters, and monitor comparisons |
+| `citable artifacts export/verify/import` | implemented — portable, versioned, checksum-bound run interchange without hosted-service authority |
 
 Shared refusal conditions (all commands): any action that would fabricate
 facts, citations, evidence, authorship, reviews, corroboration, or hidden

package/dist/universal/.agents/skills/citable/commands/artifact-interchange.md

@@ -0,0 +1,28 @@
+# Portable artifact interchange
+
+Use artifact interchange to move one sealed Citable run between local or
+external workspaces without changing its canonical bytes.
+
+```sh
+citable artifacts export <run-id> --output <directory>
+citable artifacts verify --input <directory>
+citable artifacts import --input <directory>
+```
+
+An export contains the original run package plus
+`citable-artifact.json`. The versioned envelope binds the run identity,
+exporter identity, source-run versions, manifest and checksum hashes, exact
+artifact inventory, byte sizes, and a canonical package hash. Verification rejects missing,
+additional, modified, unsafe, or symbolic-link entries. Import verifies into a
+temporary directory before atomically creating `.citable/runs/<run-id>`.
+
+Import never rewrites the run manifest, checksums, or artifacts. An identical
+existing run is `already_present`; a different package with the same run ID is
+a collision and fails closed. The interchange envelope is transport metadata
+and is not copied into the canonical run package.
+
+Checksums prove preservation after export. They do not prove that the source
+was authentic, that external observations were representative or complete, or
+that an intervention caused an outcome. Hosted tenancy, authentication,
+storage, retention, billing, scheduling, and dashboards are outside this
+contract.

package/dist/universal/.agents/skills/citable/commands/field-validation.md

@@ -0,0 +1,71 @@
+# Field validation corpus
+
+Use this workflow to assemble, evaluate, and publish an acceptance corpus. A
+fixture-valid corpus is not field evidence. Do not describe a corpus as real,
+representative, accurate, or public until the corresponding records and reviews
+exist.
+
+## Preconditions
+
+For every property, record:
+
+- architecture and property class;
+- owner authorization for collection, retention, review, and the intended
+  publication level;
+- collection methods, authority labels, timestamps, and limitations;
+- unavailable evidence, intentional incompleteness, and contradictions;
+- publication approval, retention, and the exact artifact-reference allowlist;
+- human sanitization review covering personal data, credentials, and source
+  identifiers.
+
+Every execution must reference raw artifacts, deterministic and heuristic
+findings, reviewer assignments, disagreements, exceptions, interventions,
+verification, and its reproducibility receipt where available. Empty reference
+arrays are valid only where that class of evidence did not occur.
+
+## Evaluate
+
+```sh
+citable corpus evaluate --input acceptance-corpus.json
+```
+
+Evaluation validates the v2 contract and writes `accuracy-metrics.json` plus a
+human-readable `accuracy-metrics.md` projection from the same hash-bound data.
+Every reported metric discloses its numerator, denominator, population,
+exclusions, and confidence limits. Sample and census designs remain distinct,
+and detector versions are reported as separate cohorts. Evaluation does not
+authorize publication. Unknown false negatives remain unknown; discovered
+false-negative rates apply only to adjudicated expected detections in the
+disclosed population.
+
+## Publish
+
+```sh
+citable corpus publish \
+  --input acceptance-corpus.json \
+  --output public-corpus.json
+```
+
+Publication fails closed when any property is private or aggregate-only, owner
+authority is expired or incomplete, sanitization review is incomplete, an
+artifact reference is unsafe or outside the allowlist, sensitive patterns are
+detected, or the output already exists. It emits the public corpus, a hash-bound
+publisher-controlled receipt, and an immutable local evidence package.
+
+Automated checks supplement but do not replace human sanitization review. The
+receipt is owner-authorized publication evidence, not independent attestation.
+Never publish production credentials, customer data, private property artifacts,
+or evidence outside the exact approved scope.
+
+## Reproducibility
+
+Create one receipt per sealed acceptance run, then compare repeated envelopes:
+
+```sh
+citable corpus receipt --run <run-id> --input execution-context.json
+citable corpus compare-receipts <receipt-a.json> <receipt-b.json>
+```
+
+Treat fingerprint equality as bounded equality of declared inputs, methods,
+tools, external-system versions, and artifacts. It does not establish causation
+or performance on untested properties.

package/dist/universal/.agents/skills/citable/commands/observe-and-act.md

@@ -9,7 +9,16 @@ failure_behaviour: unavailable dependency, credential, source run, reviewer, exa
 ## Collection modes
 
 - `observe render --target <URL>` uses optional Playwright. Browser evidence is
-  not search-index evidence.
+  not search-index evidence. It captures independent desktop, mobile, and
+  JavaScript-disabled profiles. `--interactions` exercises at most 20 visible
+  disclosure, inactive-tab, and load-more-like controls; it does not prove an
+  application-specific journey. `--resume-run` reuses only successful profile
+  observations from the named immutable run and recollects failed/absent ones.
+  `observe render --input <browser-evidence-plan.json>` executes an
+  explicit Chromium, Firefox, or WebKit profile matrix. Each plan records the
+  browser version, device, JavaScript, locale, consent, authentication state,
+  and ordered interaction steps. Fill/select values are referenced by
+  environment-variable name and are never written to the plan or artifacts.
 - `observe index --input <JSON>` imports owner exports. With `--target`,
   `--site-url`, and `GSC_ACCESS_TOKEN`, it calls Google URL Inspection, which
   reports the indexed version rather than performing a live indexability test.
@@ -21,6 +30,9 @@ failure_behaviour: unavailable dependency, credential, source run, reviewer, exa
   without IP verification remain low confidence.
 - `observe passages|consensus --target <dir|URL>` performs static extraction.
 - `observe performance` imports evidence or calls CrUX with `CRUX_API_KEY`.
+  `--lighthouse` instead performs 1-5 local lab runs (default 3) using optional
+  `lighthouse` and `chrome-launcher` peers, preserving each result and a median
+  summary. Lab evidence remains separate from CrUX field evidence.
 - `observe corroboration --input <JSON>` records observable external mentions;
   controlled or unverified sources do not become authority claims.
 

package/dist/universal/.agents/skills/citable/manifest.json

@@ -1,29 +1,31 @@
 {
   "name": "citable",
-  "version": "1.11.0",
+  "version": "1.13.0",
   "managedBy": "citable-cli",
   "provider": "codex",
   "providerName": "OpenAI Codex",
   "scope": "bundle",
-  "sourcePackage": "@nebulacomponents/citable@1.11.0",
+  "sourcePackage": "@nebulacomponents/citable@1.13.0",
   "variantPath": ".agents/skills/citable",
   "files": {
     "commands/action-findings.md": "sha256:510361461f24fcd4a550e0a8d960c383aefe23f9a65a54a3ffc44abd4c1293e7",
+    "commands/artifact-interchange.md": "sha256:18d56fbfa6bb3f23f08593a9a8135e51c268618b049e53c793969b68000875c3",
     "commands/delivery.md": "sha256:3b29250f9e9f84442812e320a70d5d2784b4286e91af3866f99cb6062a85a10f",
+    "commands/field-validation.md": "sha256:4f24b2758eef746d704b218364f1e45ee9e07418c8ae111081b1fd41f93dde2d",
     "commands/governance.md": "sha256:1fd6272f18b00237fc6e602ac126ae812747937b2bb575068a0d617137eea2b0",
     "commands/ingest.md": "sha256:cf35659aecd9d0e160909003efdbe056864fc5d824c71b4ef0ee801ce07ff598",
     "commands/measurement.md": "sha256:9cade6dfec23338e18aa67ede3335ac9e2888dc577a7d4d8e3ceac948bce403f",
     "commands/media-evidence.md": "sha256:f2218feabe3360a2c9d569a5fa9b70d4b33a83a270d657151635e6f404fb34e0",
-    "commands/observe-and-act.md": "sha256:02a93bbbfddc3d8e11cb921ddccd19b0666133c4bdb4f4a280bcb131bed37e8c",
+    "commands/observe-and-act.md": "sha256:3862708da9d787d61a9c7bc8639b2fb28bc021d5784e11f313b0e9ecdff261bc",
     "commands/optimize-page.md": "sha256:35866f3d7cdeef0376ea7a7d7036085a9b9cf95bbb300ecf4bd77610ebd74d85",
     "commands/page-work.md": "sha256:965a66e59bdf7b0da482d26fade53b3d265885ece5d463fe83b0789ebefaef3e",
-    "commands/README.md": "sha256:36a0462ce9867dd46922185060341110f9c4b71582b96ec3f90ff1ff341f59ea",
+    "commands/README.md": "sha256:c27f69e0e33712e00f780c0988c7b033650c17d3450baea33f2b14212be8438e",
     "commands/reviews.md": "sha256:902904aaa405fe5724cf0662c7a5a36eab9a4fe0eb81cbd95ffa2fff946e0ac8",
     "commands/validate-render.md": "sha256:a8e1d04d40b84d57c35039250d98c6a7302cd44a148a8c3b9d9cfb92678c4347",
     "policies/crawler-policy-template.md": "sha256:15c88a4eeb21f53f71e5250dd3eb2d2791a814b304eaefce8bb30e2bdf2e2e59",
-    "references/aeo-geo-validation.md": "sha256:e45828d3ccca54a08513cb4235578c8a8ecf1dc47d1cb98e4a497838103bef70",
+    "references/aeo-geo-validation.md": "sha256:4b0041764f2fdbfbff1e1a2907eaed02cf0aef491254a9c98ec9250ff1084e80",
     "references/anti-patterns.md": "sha256:de63f8e646af0fbf479b45eb1d050ae62a8637acc320d21b87e4a083fd4c47e7",
-    "references/capability-boundaries.md": "sha256:5d59f15cf97529fec26a105095278dc7a17ae3b829103994ec5ddf420897f448",
+    "references/capability-boundaries.md": "sha256:4901a32efbe747c8fed54ed1bffe345603c03a1bf5906b7f6da91395720f00e0",
     "references/evidence-authority.md": "sha256:b81ee09953999f16b19dc0eaba6fffe794b58d152e2adbd95c3179730bc08ffc",
     "rubrics/answer-extractability.md": "sha256:c8e741143c34e53afdb3dc7b6976ce7b4485dce13b12d14c193fccb17bacf1da",
     "rubrics/claim-boundedness.md": "sha256:0bab1b85248c252b966175912ebfae13bf18aa7e41573f7466831ac78b5d3e27",
@@ -38,43 +40,61 @@
     "rubrics/recommendation-eligibility.md": "sha256:38d2055fc5f6ab5e40fd1b48de4e0518fc4382da0b7062ad09c5afebcc673fc3",
     "rubrics/semantic-clarity.md": "sha256:66d9f4843a4a620ea8c00625c5e17c02baeaa373f2b445ce13de6f81d78ec7ba",
     "rubrics/source-authority.md": "sha256:8dbfe46bdcd0b5351f038be503601268c7d0abb9bed83be9ca58bf54dec5a73b",
+    "schemas/acceptance-corpus.schema.json": "sha256:b703be371ee33747d83814e448d18cd01f103f1832480a95df4331b86982165d",
+    "schemas/acceptance-run-context.schema.json": "sha256:56d75c7bcd166c3578a71699df1aa5466327c23c1e06e33773f2493f5caa1a50",
+    "schemas/acceptance-run-receipt.schema.json": "sha256:6c6e66767cf3b9adb42ace087f7e9a428312f41b7ecbaf533fc8d7ba5953da44",
     "schemas/action-plan.schema.json": "sha256:4b6c4c877840a065c7c0c0c7b52f696502bf3d4ca6d31179622fc364ea868d5e",
+    "schemas/artifact-interchange.schema.json": "sha256:c6f628934c8a1b8ea2996e03c2c5abd46c2a0e74c680b0a43e44c2343dc9aa21",
+    "schemas/assumption.schema.json": "sha256:31a0f44bbf1d207e013961ec4c39eea85c3ae9efde1485b5e26084562c816844",
     "schemas/bing-webmaster.schema.json": "sha256:8357c518a8cb4ccd1def2da38950fbba68270e2b1613ad522b50c568a3b8538f",
+    "schemas/browser-evidence-plan.schema.json": "sha256:f2c1a3fe05cdaf72a41c0a3d548ccf5bdf247dbe71b546759b15277df0e723ba",
     "schemas/claim.schema.json": "sha256:ce79ccbfa34ece9b7a309a6962522ea06a1f44c468b4c472b5c376554a788537",
     "schemas/common.defs.json": "sha256:effdd812910d1c275c50520bb70d8aec085e53fa8cb68054dc58c04a5d4a66b3",
     "schemas/competitor.schema.json": "sha256:2f6fd7e4543846bde56ff4dc62bb41fa366889787b794a7cdb0bff1a3392e591",
     "schemas/config.schema.json": "sha256:a87f3bb0087de09035b9f2dafea66e32e4fcc2e3e329abcd0fd8a5ec33318f2a",
     "schemas/connection.schema.json": "sha256:abcedde8d62d0c0663d4f959e06132a617908058ed965078965445ef3b108310",
+    "schemas/corpus-publication-receipt.schema.json": "sha256:a4d523fd1c1fa53762522891c06c06edb668cd15e1de3c7123d2b9948445ae32",
     "schemas/crawler-identity.schema.json": "sha256:e1e3e6da203a8bc49fba8f9bbde525368891db022d542fc615b677401825351d",
     "schemas/crawler.schema.json": "sha256:1a113dfb80bc6cde904c3ea648c82ea1d88fe50f2c1a0e7175d44abba28ac64e",
+    "schemas/customer-outcome.schema.json": "sha256:c0cdfee31a692feadb31dc6be734d45ae178ee53f28349c2a9e7504bdfcf9e50",
+    "schemas/decision.schema.json": "sha256:7c30f3ad83a6aa0b67ecc51ac6691e5e041b633d01bab17cee2b5218d926204e",
+    "schemas/deployment-receipt.schema.json": "sha256:789455907e7f3039fd55ee5b2568e09871128f513f5ddfe1f3afac1a4dd6b2bd",
     "schemas/detector-quality.schema.json": "sha256:5e9e1ac80b1198a0ff8fa17b9f7a543d3ccf3ed3d60196152f6ff095cfdebccd",
     "schemas/entity.schema.json": "sha256:78a08c35c8f4eb7a804ab46b61a238502f59621ac5218991e783251fbde61c4f",
     "schemas/evidence-authority.schema.json": "sha256:4f8165a37304ebab019e1c2dfec13f1168ecf89ce9bec88f7dca9c8554e39ac0",
     "schemas/evidence.schema.json": "sha256:76baec7ad7e89c6952473c01dacb9e3b26953c088331e6314d531ce620df9ec1",
     "schemas/exception.schema.json": "sha256:c88985494ec66554330fdc5b65d033527d0005acf41f10798b1e831d171f8944",
     "schemas/experiment.schema.json": "sha256:3cebebc5745a9eb359261408b0a0421b8370d2c94f8244ce939827cd9da41cb6",
+    "schemas/field-validation-metrics.schema.json": "sha256:bd5daeafc1593f1c4707d75180e1feb1e81d9345a57c1f269eadf13c36cb7012",
     "schemas/finding.schema.json": "sha256:85907cd04cd8d2fe24a1e618e0df7160f6d56aa1818db501f345aebce5eb5edb",
+    "schemas/initiative.schema.json": "sha256:491f6cc022c7ac1f9ea9ac9ddfcd270cb49c390e2601fe1e73aa6192a9e7689a",
     "schemas/intervention.schema.json": "sha256:fcf28e117b207e64ffc96088aa70a95370841a2a9fa43ecb1c82b34c2058a818",
+    "schemas/kpi.schema.json": "sha256:a45e2178ac03a97692b2c1dd667abde6b25b936c73ba0c32e4e98becb3d9d1c2",
     "schemas/media-manifest.schema.json": "sha256:3f37301aab9e913fa7febb15c720ed6384b106f12cef8d26cdd12d93fd85b41f",
     "schemas/metric.schema.json": "sha256:2ada648e7274a8d78cc9d6e7ed616f91faa79cf9ed07613e2c1fc53b247aac8d",
     "schemas/objective.schema.json": "sha256:94404dc95d04a6641ea7b9654317d0222793251ad44d3dfd2875738622a36025",
-    "schemas/observation.schema.json": "sha256:eebb7deb445b9de8c2ffa603fce441f9a857ce290602d8828915fb630b566693",
+    "schemas/observation.schema.json": "sha256:b572ae8ba3d175509cc9fae914d8d054c0fb84ebcaf81de981532b527064538d",
     "schemas/page.schema.json": "sha256:627613c35f461177fa946df08edf66fbf6f6bbdfc57eef812572b3af796d5e07",
     "schemas/project.schema.json": "sha256:ea48fdb6480f7047ba4594bd5908bc1f79c1ad6b9a2c204df900959b5f203dc6",
     "schemas/prompt-result.schema.json": "sha256:b55be47d7d58d7b1dd8c3b1e849e24e55216212c592cde7eb836953379b85737",
     "schemas/prompt.schema.json": "sha256:7f0f50c83a86ac91faa7b79895dbc1fed371e8c0342e18d8c2877cc81f4f0f4c",
     "schemas/query.schema.json": "sha256:3c0793a8c0c48ab39320416790840b345ea987696f35b1cc364f23245a62aacb",
+    "schemas/release-manifest.schema.json": "sha256:962be645e790b5f3dc6f6ea455e187492d727dcb91d00f91e4e25b40e2dbc360",
+    "schemas/release-state.schema.json": "sha256:9920f68e538915d0d07044f313a9e410775a38ce9bb46a55d9ebe02970784803",
     "schemas/remediation-spec.schema.json": "sha256:6fe8cb773702862ba6d261198bb5cc2dca3eca0a2310a60ebddd04377654df8d",
     "schemas/review-item.schema.json": "sha256:df532a70385d5031bbc18654f3ccb436dc6e8d402ccf987e58f8f7e3ad2fb3f1",
     "schemas/review-policy.schema.json": "sha256:b2d176b1886f78c384b25532a0c3eb18548dfbd17f362a305e622bab562a1d04",
     "schemas/reviewer.schema.json": "sha256:3716697b752575a1277f2c2b318b1f2565de02bf9dce2a0c165567ab5330d18b",
+    "schemas/risk.schema.json": "sha256:ea3ca47a8457d98ea65df78e07ba6c20479906e8a3df07a6fa4be5644828fc49",
     "schemas/run.schema.json": "sha256:46c3178620f3a358d936d535aeadb83454b9b35b551e53e6615cc9862063dacb",
     "schemas/sampling-plan.schema.json": "sha256:005f70f149baaacd7de698548d556d3e401c630978748271b0e14a74726dd413",
+    "schemas/scenario.schema.json": "sha256:3f5bc2032d08f973600db068f475176a0e4706b33be42f99c778fe755a0ff38b",
     "schemas/schedule.schema.json": "sha256:0f02174ab0fc7bbbce9759fd358b18ed97cc683c76933fd3001cf053c499b157",
+    "schemas/variance.schema.json": "sha256:68389f64f0a3be13ff8951e8eefba9a5c31d0738325cb47e37c1a8435bea48f0",
     "scripts/README.md": "sha256:1c3e952950ef353849621b07c5f615bf7cd20cc5654da4635492a7bd6384e747",
-    "SKILL.md": "sha256:d07fdc4d4cbc11b5c530bd97bf3e4d13d9c212cc3c67cf9798042f776f68609c",
+    "SKILL.md": "sha256:ac5817aa956c0d845ab4cc3adcc07b8bd814f924883b9c1178cbe745333616d1",
     "templates/finding-review.md": "sha256:b270a83d3f015150aac80249ae9314979c0befea39ff92ec455f9a5aaa73404c",
-    "VERSION": "sha256:74ee4d121463989183f9a9b848577eb2434f3f5d095de2a4b197bf29ffa2399b"
+    "VERSION": "sha256:b4a9545a385e9701482be56a0efee678d467d9c586ec476f2c5986cfaa24e96c"
   },
-  "treeHash": "sha256:26c75b5cf68676c665ecf6a1e8d4ec5c6e77abdef8825c1c3ad8a099147168b8"
+  "treeHash": "sha256:8c0cfbcf3b6ed8a20ec337f655a3ee299f3f7cfd837bad5f7beb8627f3954242"
 }

package/dist/universal/.agents/skills/citable/references/aeo-geo-validation.md

@@ -9,7 +9,7 @@ not a prediction of citation or recommendation. Report every row as
 | State | Current evidence boundary |
 | --- | --- |
 | Retrieval eligibility | Static policy plus captured synthetic HTTP response; crawler identity remains simulated unless verified logs/IP evidence are supplied |
-| Source extraction and support suitability | Initial-HTML extraction plus registries/detectors/rubrics; rendered-DOM parity remains not established without a renderer |
+| Source extraction and support suitability | Initial HTML plus registries/detectors/rubrics; optional fixed Chromium profiles and explicit Chromium/Firefox/WebKit journey plans establish bounded artifact-level parity, while semantic impact remains unproven |
 | Observed citation behavior | Operator-supplied prompt-result observations only; report cohort size and property citation presence, never infer from readiness |
 
 ## AEO checks

package/dist/universal/.agents/skills/citable/references/capability-boundaries.md

@@ -9,6 +9,9 @@ Use the narrowest supported state. Never promote evidence across rows.
 | Verified crawler reached edge | Imported evidence | `observe logs` record plus provider-published IP verification where supported |
 | Observed in production logs | Imported evidence | Timestamped `observe logs` artifact with URL, status, bytes, latency, cache result, UA, and verification result |
 | Indexed | Google API or imported evidence | `observe index`; Google inspection covers the indexed version, other engines require owner exports |
+| Rendered profile observed | Optional local browser evidence | Successful profile-specific Chromium DOM, text, screenshot, failed requests, interactions, viewport, and configuration; failed profiles remain independent |
+| Browser journey observed | Schema-validated local browser plan | Engine/version, device, JavaScript, locale, consent, authentication declaration, steps, DOM, text, accessibility snapshot, screenshots, and failures remain separate; differences do not establish semantic or citation impact |
+| Render parity compared | Bounded local comparison | Raw/desktop and mobile/desktop word ratios from the same run or explicitly resumed immutable profiles; not semantic equivalence |
 | Returned by retrieval | Operator supplied | Provider observation containing retrieved source URL |
 | Cited | Import or disclosed adapter | Complete `observe citations` answer with citation URL and product/adapter mode |
 | Materially supports answer | Human review required | `citation_review` maps answer claim to source passage and names the reviewer |