Enhance ControlFlow Orchestration and Planning Skills

- Updated SKILL.md in controlflow-orchestration to refine execution guidelines, including the introduction of Pre-Wave Cache Guard and improved phase serialization.
- Added new references for decision challenges and source grounding to support high-risk decision-making and external claims validation.
- Enhanced failure taxonomy with a diagnosis packet for retries and introduced transient wave throttling for improved error handling.
- Revised phase checklist to include inspections for shared components and recent task artifacts.
- Updated runtime policy to include new parameters for transient reduction and batch approval.
- Improved planning skills with requirements for revision lineage and verified item checks.
- Added negative test fixtures to validate strict plan lifecycle and tier review requirements.
- Introduced core portability matrix to document selective core-to-plugin portability contracts.

Author: Smith

NOTES.md

@@ -2,6 +2,6 @@
 
 Repo-persistent active-objective state only. See `docs/agent-engineering/MEMORY-ARCHITECTURE.md` for the three-layer memory model; task-episodic history lives under `plans/artifacts/<task-slug>/`.
 
-- Active objective: Auto-default + selective pinning complete (4 pinned agents); full eval gate (`cd evals && npm test`) green.
-- Blockers: none.
-- Pending: optional follow-up to investigate pre-existing capability-matrix live-tree drift flags surfaced by the smoke test.
+- Active objective: Codex plugin selective-parity audit complete; LARGE implementation plan ready for PLAN_REVIEW.
+- Blockers: Full eval baseline fails at the pre-existing Pass 10 cross-plan overlap gate.
+- Pending: Resolve the overlap baseline through its owning plan before final plugin-parity verification.

Planner.agent.md

@@ -49,6 +49,8 @@ For all other scopes, record applicability, impact, evidence source, and disposi
   2. Match task keywords and domain signals against the index.
   3. Select ≤3 most relevant skill files based on task context and complexity tier.
   - For spec-bearing phases, consider `skills/patterns/spec-driven-development.md` when requirements, scope boundaries, acceptance criteria, or spec-before-plan discipline are central to safe planning.
+  - For phases depending on external framework/API/version claims or sibling-source adoption, consider `skills/patterns/source-grounding.md`.
+  - For high-risk or non-trivial decisions with materially different alternatives, consider one bounded `skills/patterns/decision-challenge.md`.
   4. Include selected skill file paths in each applicable phase's `skill_references` array.
   Implementation agents load referenced skills before executing phase tasks.
 6. Research (delegate CodeMapper-subagent/Researcher-subagent when scope is large).
@@ -115,6 +117,8 @@ Agent-specific additions:
 - `plans/project-context.md` (if present)
 - `skills/index.md` (domain skill mapping — read during Step 5)
 - `skills/patterns/llm-behavior-guidelines.md` (load on non-trivial tasks — anti-pattern guardrails: scope drift, over-abstraction, silent assumptions, weak success criteria)
+- `skills/patterns/source-grounding.md` (consider for external claims, provenance, and explicit `UNVERIFIED` state)
+- `skills/patterns/decision-challenge.md` (consider once for high-risk or non-trivial decisions)
 - Plan artifacts directory: `plans/` (default location for all plan and completion files)
 
 ## Tools

README.md

@@ -265,7 +265,7 @@ See [`evals/README.md`](evals/README.md) for pass descriptions and how to add sc
 │   ├── tutorial-en/               # Full English-language tutorial (19 chapters)
 │   └── tutorial-ru/               # Full Russian-language tutorial (19 chapters)
 ├── governance/                    # Operational knobs and tool grants
-├── skills/                        # Reusable domain pattern library (18 patterns)
+├── skills/                        # Reusable domain pattern library (20 patterns)
 ├── evals/                         # Offline validation suite
 │   └── scenarios/                 # Eval scenario fixtures
 ├── plans/                         # Plan artifacts and templates
@@ -324,6 +324,8 @@ A portable adaptation of ControlFlow for [OpenAI Codex CLI](https://github.com/o
 
 The plugin brings the core ControlFlow disciplines — phased planning, pre-execution plan review, assumption verification, orchestration, evidence-backed code review, and memory hygiene — into Codex without depending on VS Code-specific tool contracts, fixed agent rosters, or `@Agent` syntax.
 
+Version `0.6.0` adds machine-checked selective parity through `plugins/controlflow-shared-source/core-portability-matrix.json`: portable workflow invariants are adopted or adapted, while model routing, tool grants, the fixed roster, session telemetry, compaction, budgets, and `model_unavailable` remain explicit divergences.
+
 ### Included Skills
 
 | Skill | Analogous ControlFlow Role |
@@ -449,11 +451,10 @@ See [`plugins/controlflow-codex/USAGE.md`](plugins/controlflow-codex/USAGE.md) f
 powershell -ExecutionPolicy Bypass -File plugins/controlflow-codex/scripts/validate-strict-artifacts.ps1 `
   -RepoRoot . `
   -PlanPath plans/my-task-plan.md `
-  -RequirePlanAudit `
-  -RequireAssumptionVerifier
+  -StrictReviewByTier
 ```
 
-The `validate-strict-artifacts.ps1` script validates **Codex strict-plan artifacts only**. Do not use it for core VS Code plans. It enforces the mandatory lifecycle sections (`## Progress`, `## Discoveries`, `## Decision Log`, `## Outcomes`, `## Idempotence & Recovery`).
+The `validate-strict-artifacts.ps1` script validates **Codex strict-plan artifacts only**. Do not use it for core VS Code plans. It enforces the mandatory lifecycle sections (`## Progress`, `## Discoveries`, `## Decision Log`, `## Outcomes`, `## Idempotence & Recovery`) in exact order. `-StrictReviewByTier` derives required review artifacts from tier and applicable unresolved `HIGH` risk; the existing `-Require*` switches remain additive compatibility controls.
 
 ### Intentional Differences from the VS Code Version
 

Researcher-subagent.agent.md

@@ -31,6 +31,7 @@ If `context_packet` is present in your dispatch, read the referenced `artifact_p
 - Every claim requires evidence (`file`, `line_start`, optional `line_end`).
 - If evidence is insufficient, output `ABSTAIN`.
 - Separate observed facts from hypotheses explicitly; tolerate naming/format variance without speculative inference.
+- For external framework/API/version claims, follow `skills/patterns/source-grounding.md`: prefer current official primary sources and label material unsupported claims `UNVERIFIED`.
 
 ## Archive
 
@@ -55,6 +56,7 @@ See [skills/patterns/preflect-core.md](skills/patterns/preflect-core.md) for the
 - `schemas/code-mapper.discovery.schema.json`
 - `docs/agent-engineering/PROMPT-BEHAVIOR-CONTRACT.md`
 - `plans/project-context.md` (if present)
+- `skills/patterns/source-grounding.md`
 
 ## Tools
 

docs/tutorial-en/00-introduction.md

@@ -27,7 +27,7 @@ The repository delivers:
 | Agent prompt files (`*.agent.md`) | 13 | Define role, scope, contracts, and tools for each agent |
 | JSON schemas (`schemas/*.json`) | 17 | Formal contracts for inter-agent communication |
 | Governance configs (`governance/*.json`) | 7 | Tool permissions, retry budgets, tier routing, model mapping |
-| Skill patterns (`skills/patterns/*.md`) | 18 | Reusable domain expertise loaded just-in-time |
+| Skill patterns (`skills/patterns/*.md`) | 20 | Reusable domain expertise loaded just-in-time |
 | Eval harness (`evals/`) | ~410 checks | Offline validation of schema compliance and behavioral invariants |
 | Documentation (`docs/`) | — | Engineering policies, tutorials, architecture docs |
 

docs/tutorial-en/11-skills.md

@@ -75,7 +75,7 @@ sequenceDiagram
     E->>E: apply patterns
 ```
 
-**Why just in time?** Skills add context to the agent's prompt. Loading all 18 skills upfront wastes tokens and creates noise. Only load what is needed for the current phase.
+**Why just in time?** Skills add context to the agent's prompt. Loading all 20 skills upfront wastes tokens and creates noise. Only load what is needed for the current phase.
 
 ## Key Skills Deep Dive
 

docs/tutorial-en/17-glossary.md

@@ -54,7 +54,7 @@ Alphabetical reference for all key terms. Each entry includes a definition and c
 
 **Confidence** — A numeric value (0–1) in Planner output and gate events reflecting how certain the agent is in its assessment. → Ch.06
 
-**ControlFlow** — A prompt/governance/eval repository for an agent orchestration system. 13 agents, 17 schemas, 7 governance files, 18 skills. → Ch.00
+**ControlFlow** — A prompt/governance/eval repository for an agent orchestration system. 13 agents, 17 schemas, 7 governance files, 20 skills. → Ch.00
 
 **Convergence detection** — See stagnation. → Ch.07
 

docs/tutorial-ru/01-quickstart.md

@@ -20,7 +20,7 @@ ControlFlow/
 │   └── project-context-registry.json
 ├── skills/
 │   ├── index.md                     # каталог skill-паттернов
-│   └── patterns/                    # 18 паттернов
+│   └── patterns/                    # 20 паттернов
 ├── evals/                           # оффлайн-проверки
 │   ├── package.json                 # npm test → канонический gate
 │   ├── validate.mjs                 # структурный проход

docs/tutorial-ru/README.md

@@ -33,7 +33,7 @@
 | 08 | [Пайплайн исполнения](08-execution-pipeline.md) | Волны параллельного исполнения, executor_agent, quality gates, completion gate, optional final review | 8 |
 | 09 | [Схемы (контракты)](09-schemas.md) | Разбор всех 17 JSON-схем: какая для какого агента, ключевые поля | 10 |
 | 10 | [Governance](10-governance.md) | agent-grants, tool-grants, runtime-policy, model-routing, rename-allowlist | 8 |
-| 11 | [Skills (паттерны)](11-skills.md) | Skill library, протокол выбора (≤3 на фазу), 18 паттернов | 7 |
+| 11 | [Skills (паттерны)](11-skills.md) | Skill library, протокол выбора (≤3 на фазу), 20 паттернов | 7 |
 | 12 | [Архитектура памяти](12-memory.md) | Трёхслойная модель: session / task-episodic / repo-persistent; роль NOTES.md | 6 |
 | 13 | [Таксономия сбоев](13-failure-taxonomy.md) | 4 класса: transient / fixable / needs_replan / escalate; retry-политики | 6 |
 | 14 | [Eval-харнесс](14-evals.md) | Устройство `cd evals && npm test`, режимы full/structural/behavior | 6 |

evals/README.md

@@ -179,6 +179,8 @@ npm test
 | **12 — Governance Policy Assertions** | Runtime policy invariants for tool-output spill, session telemetry, model-routing hints, compaction, memory promotion order, and CodeReviewer security mode. |
 | **13 — Final Review Coupling** | `review_scope=final` coupling between CodeReviewer prompt behavior and schema fields, including injected context expectations. |
 | **14 — Canonical Source Matrix** | Ensures `plans/project-context.md` keeps the Canonical Source Matrix heading. |
+| **15 — Tool Labels / Pattern Budget / Doc Counts / Generation Parity** | Runs the existing live-tree drift checks for registry tool-count labels, skill-pattern line budgets, allowlisted documentation totals, and no-delta Codex generation parity. |
+| **16 — Selective Plugin-Core Portability** | Validates `core-portability-matrix.json`: unique invariant IDs, allowed dispositions, evidence paths, and declared semantic anchors without requiring byte parity with core prose. |
 
 ### Separate harnesses in `npm test`