From 5d054227cd096cf2945c87c9fe59a4e89b199bcf Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=D0=94=D0=BC=D0=B8=D1=82=D1=80=D0=B8=D0=B9?= Date: Sun, 17 May 2026 18:21:49 +0300 Subject: [PATCH] =?UTF-8?q?docs(c10):=20plan=20correction=20=E2=80=94=20op?= =?UTF-8?q?erations=20is=209=20skills,=20not=20/ops:*=20commands?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Task 2 install revealed operations@knowledge-work-plugins v1.2.0 ships 9 skills (process-doc, process-optimization, change-request, …) and 0 lifecycle hooks — not /ops:* slash-commands. OPS4 resolved on install; +OPS5 (boundary vs the 2 self-authored skills); skill "Границы" sharpened. cspell-words += RACI/DMN/czlonkowski. Co-Authored-By: Claude Opus 4.7 (1M context) --- cspell-words.txt | 5 ++ ...10-business-process-tooling-integration.md | 59 +++++++++++-------- 2 files changed, 41 insertions(+), 23 deletions(-) diff --git a/cspell-words.txt b/cspell-words.txt index 2dd5691a..6104940e 100644 --- a/cspell-words.txt +++ b/cspell-words.txt @@ -1391,3 +1391,8 @@ ivotoby дровер канбаном коммитить + +# C10 business-process tooling integration — spec + plan (2026-05-17) +RACI +DMN +czlonkowski diff --git a/docs/superpowers/plans/2026-05-17-c10-business-process-tooling-integration.md b/docs/superpowers/plans/2026-05-17-c10-business-process-tooling-integration.md index b941b3a1..02fe48c8 100644 --- a/docs/superpowers/plans/2026-05-17-c10-business-process-tooling-integration.md +++ b/docs/superpowers/plans/2026-05-17-c10-business-process-tooling-integration.md @@ -16,13 +16,13 @@ | # | Tool | Install mode | Source / License | Hooks? | |---|---|---|---|---| -| 1 | **operations plugin** — business-process documentation (`/ops:process-doc`), change management (`/ops:change-request`), capacity planning (`/ops:capacity-plan`), vendor review (`/ops:vendor-review`). | Claude Code **marketplace plugin** — `operations@knowledge-work-plugins`; marketplace `anthropics/knowledge-work-plugins` (already added — #42/#46 installed from it) | GitHub `anthropics/knowledge-work-plugins`, Anthropic Verified | Verify on install (Task 2 Step 4) — knowledge-work plugins are skills + slash-commands; #42/#46 added no lifecycle hooks | +| 1 | **operations plugin** — 9 business-process skills: `process-doc`, `process-optimization`, `change-request`, `capacity-plan`, `compliance-tracking`, `risk-assessment`, `runbook`, `status-report`, `vendor-review`. | Claude Code **marketplace plugin** — `operations@knowledge-work-plugins` v1.2.0; marketplace `anthropics/knowledge-work-plugins` (already added — #42/#46 installed from it) | GitHub `anthropics/knowledge-work-plugins`, Anthropic Verified | **None** — skills-only plugin, no `hooks/` dir, no `commands/` dir (verified on install 2026-05-17, Task 2) | | 2 | **process-modeling skill** — BPMN 2.0 notation, process maps, customer-journey / value-stream maps, RACI matrices, state-machine modeling. Renders via the `mermaid` skill. | **Self-authored** standalone skill — created in `.claude/skills/process-modeling/` (no plugin, no marketplace) | Project-authored — content in this plan (Task 3) | None — self-authored, no `hooks` block | | 3 | **process-analysis skill** — process discovery (reverse-engineer the as-is process from Laravel code + audit-log tables), bottleneck analysis, requirement→process traceability, business-process KPI/metrics. | **Self-authored** standalone skill — created in `.claude/skills/process-analysis/` | Project-authored — content in this plan (Task 4) | None — self-authored, no `hooks` block | | — | **n8n-mcp** (`czlonkowski/n8n-mcp`) — workflow-automation MCP server for the n8n platform. | **NOT installed** — deferred reserved slot (see "Deferred Task"). | GitHub `czlonkowski/n8n-mcp`, **MIT** | n/a — not installed | | — | **mermaid** / **architecture-patterns** / **CCPM** / **product-management** / **superpowers writing-plans** | **Reuse** — already installed; surfaced in C10 via `NODE_SECTION_SECONDARY` | A6 #37/#38, C9 #41/#42, superpowers | n/a | -**Verification status:** operations — plugin **name `operations`** confirmed via the live `anthropics/knowledge-work-plugins` `.claude-plugin/marketplace.json` (WebFetch 2026-05-17); description "Optimize business operations — vendor management, process documentation, change management, capacity planning, and compliance tracking"; plugin-id form `operations@knowledge-work-plugins` (matches `product-management@knowledge-work-plugins` / `design@knowledge-work-plugins`). The exact `/ops:*` command list is confirmed on install (Task 2 Step 3). n8n-mcp — repo `czlonkowski/n8n-mcp`, MIT, Claude Code compatible, confirmed via WebSearch; **not installed by this plan**. +**Verification status:** operations — plugin **name `operations`** confirmed via the live `anthropics/knowledge-work-plugins` `.claude-plugin/marketplace.json` (WebFetch 2026-05-17); plugin-id `operations@knowledge-work-plugins`. **Installed 2026-05-17 (Task 2): v1.2.0, scope user.** It ships **9 skills** (`process-doc`, `process-optimization`, `change-request`, `capacity-plan`, `compliance-tracking`, `risk-assessment`, `runbook`, `status-report`, `vendor-review`) — **not** `/ops:*` slash-commands — and **no lifecycle hooks** (no `hooks/` / `commands/` dir). n8n-mcp — repo `czlonkowski/n8n-mcp`, MIT, Claude Code compatible, confirmed via WebSearch; **not installed by this plan**. **Deferred (with reason — no task in this plan):** @@ -42,10 +42,11 @@ Pattern follows the AK1–CC1 / ML1–NUM1 audits used for the A6 / C9 / D3 / A1 | # | Tool | Sev | Conflict | Resolution (locked) | |---|---|---|---|---| -| OPS1 | operations | 🟢 | `/ops:process-doc` produces a flowchart; the `mermaid` skill (A6) renders diagrams — overlapping render. | operations emits **Mermaid source**; the `mermaid` skill stays the render-SoT. operations owns the *document* + methodology, mermaid owns *rendering*. Stated in `docs/process/README.md` (Task 5) + the Tooling entry (Task 7). | -| OPS2 | operations | 🟢 | `/ops:capacity-plan` overlaps product-management `/sprint-planning` (C9) and the C8 HR domain. | Boundary by scope: operations capacity-plan = cross-functional business-process capacity; `/sprint-planning` = dev-team sprints; C8 = HR. Boundary documented in `docs/process/README.md` (Task 5). | -| OPS3 | operations | 🟢 | `/ops:change-request` overlaps adr-kit (A6) and claude-md-management. | Distinct objects: change-request = business-process change; ADR = architecture decision; claude-md-management = normative-doc edit. Documented in `docs/process/README.md` (Task 5). | -| OPS4 | operations | 🟡 | knowledge-work-plugins are built primarily for "Claude Cowork" — the plugin must verifiably activate in **Claude Code**. | Task 2 installs + Step 3 verifies the `/ops:*` slash-commands appear in-session. Strong precedent: #42 product-management and #46 design from the **same marketplace** are already integrated and work in Claude Code. If `/ops:*` do not surface → stop, report. | +| OPS1 | operations | 🟢 | The operations `process-doc` skill produces a flowchart; the `mermaid` skill (A6) renders diagrams — overlapping render. | operations emits **Mermaid source**; the `mermaid` skill stays the render-SoT. operations owns the *document* + methodology, mermaid owns *rendering*. Stated in `docs/process/README.md` (Task 5) + the Tooling entry (Task 7). | +| OPS2 | operations | 🟢 | The operations `capacity-plan` skill overlaps product-management `/sprint-planning` (C9) and the C8 HR domain. | Boundary by scope: operations `capacity-plan` = cross-functional business-process capacity; `/sprint-planning` = dev-team sprints; C8 = HR. Boundary documented in `docs/process/README.md` (Task 5). | +| OPS3 | operations | 🟢 | The operations `change-request` skill overlaps adr-kit (A6) and claude-md-management. | Distinct objects: `change-request` = business-process change; ADR = architecture decision; claude-md-management = normative-doc edit. Documented in `docs/process/README.md` (Task 5). | +| OPS4 | operations | 🟢 | knowledge-work-plugins are built primarily for "Claude Cowork" — the plugin must verifiably activate in **Claude Code**. | **Resolved on install (Task 2):** operations v1.2.0 installed (scope user), present in `~/.claude/settings.json` `enabledPlugins`, ships 9 skills auto-discovered by Claude Code, **0 lifecycle hooks**. Same marketplace as the working #42/#46. | +| OPS5 | operations vs the 2 self-authored skills | 🟢 | operations ships `process-doc` (overlaps `process-modeling`) and `process-optimization` (overlaps `process-analysis`). | Boundary: operations skills are **generic, stack-agnostic** business-process methodology. The self-authored skills cover what operations structurally cannot: **`process-modeling`** = formal BPMN 2.0 / state-machine notation rendered via `mermaid`, grounded in `db/schema.sql` (the 14-status funnel); **`process-analysis`** = as-is discovery grounded in the Лидерра codebase (`routes/`, `app/Jobs`, `audit_*` tables) — operations cannot read the repo. Each self-authored `SKILL.md` "Границы" names the operations skills explicitly; generic optimization methodology is delegated to operations `process-optimization`. | | BPMN1 | process-modeling | 🟢 | Overlap with the `mermaid` skill. | `mermaid` renders flowchart/state; it does **not** carry BPMN 2.0 semantics (pools/lanes/gateways/events methodology). `process-modeling` owns BPMN methodology and **delegates rendering** to `mermaid`. Stated in the skill's `SKILL.md` "Границы" section (Task 3). | | LINT1 | process-modeling, process-analysis | 🟡 | Self-authored skills `.claude/skills/process-{modeling,analysis}/**/*.md` are caught by the cspell + markdownlint pre-commit jobs. | **Decision:** these are **self-authored project skills** — they are **linted** like `audit-portal` / `regression` (NOT lint-ignored like the vendored `mermaid`/`ccpm`/`data-scientist`). Skill content is written lint-clean; new technical vocabulary (`BPMN`, `swimlane`, `gateway`, `RACI`, `throughput`, …) goes to `cspell-words.txt`. Task 1 Step 5 confirms project skills are not in `cspell.json` `ignorePaths` / `.markdownlintignore`; if they unexpectedly are → follow suit and record. | | PA1 | process-analysis | 🟢 | Overlap with `perf-analyzer` / `analysis:bottleneck-detect` and product-management `/metrics-review`. | Different objects: `process-analysis` = *business-process* discovery/optimization; `perf-analyzer` / `bottleneck-detect` = *code/runtime* performance; `/metrics-review` = *product* metrics. Boundary in the skill's "Границы" section (Task 4) + `docs/process/README.md` (Task 5). | @@ -64,7 +65,7 @@ Pattern follows the AK1–CC1 / ML1–NUM1 audits used for the A6 / C9 / D3 / A1 | File | Created / Modified | Responsibility | |---|---|---| | `docs/process/` | Create dir | C10 home — the business-process playbook | -| `docs/process/README.md` | Create | The BP convention: tool boundaries (operations = *document/change/capacity* · process-modeling = *model* · process-analysis = *discover & optimise* · n8n-mcp = *deferred engine*); the reuse-layer cross-ref map; the OPS1/OPS2/OPS3/PA1 boundaries | +| `docs/process/README.md` | Create | The BP convention: tool boundaries (operations = *document/change/capacity* · process-modeling = *model* · process-analysis = *discover & optimize* · n8n-mcp = *deferred engine*); the reuse-layer cross-ref map; the OPS1/OPS2/OPS3/PA1 boundaries | | `docs/process/examples/deal-lifecycle-process.md` | Create | One worked example — the deal-lifecycle process modeled as a BPMN-style swimlane (the seed artifact, the A11 `promptfoo-example` pattern) | | `.claude/skills/process-modeling/SKILL.md` | Create (self-authored) | The process-modeling skill — frontmatter + methodology | | `.claude/skills/process-modeling/references/bpmn.md` | Create (self-authored) | BPMN 2.0 element reference + the BPMN→mermaid rendering map | @@ -140,6 +141,7 @@ ls docs/adr/ ``` Read `docs/Tooling_v8_3.md` Прил. Н §0 — record the **live** tool counter and the last `§4.x` subsection number. Expected (post-A11): counter `50`, last subsection `§4.25`. Record: + - C10 numbers = `counter+1 .. counter+4`: operations, process-modeling, process-analysis, n8n-mcp (pending). - C10 subsections = next four after the last `§4.x`. - ADR number = next free after the highest existing `ADR-NNN` (expected `ADR-008` — `ADR-007-ml-ai-tooling.md` exists). @@ -188,6 +190,7 @@ Confirm `operations@knowledge-work-plugins` is listed in `~/.claude/settings.jso ## Task 3: Author the process-modeling skill (BPMN1 / LINT1) **Files:** + - Create: `.claude/skills/process-modeling/SKILL.md` - Create: `.claude/skills/process-modeling/references/bpmn.md` @@ -252,8 +255,8 @@ description: Моделирование бизнес-процесса — BPMN 2 рисует — отдаёт его mermaid. - **DDD-границы доменных процессов** — скил `architecture-patterns` (bounded context = граница бизнес-процесса). -- **Документ процесса, change-request** — плагин `operations` - (`/ops:process-doc`, `/ops:change-request`). +- **Документ процесса, change-request, оптимизация** — плагин `operations` + (скилы `process-doc`, `change-request`, `process-optimization`). - **Анализ as-is процесса** (discovery, узкие места) — скил `process-analysis`. - Этот скил — про проектирование **to-be модели**, не про вскрытие as-is. ``` @@ -348,6 +351,7 @@ git commit -m "feat(c10): add self-authored process-modeling skill (BPMN/process ## Task 4: Author the process-analysis skill (PA1 / LINT1) **Files:** + - Create: `.claude/skills/process-analysis/SKILL.md` - Create: `.claude/skills/process-analysis/references/discovery.md` @@ -414,7 +418,10 @@ description: Анализ и оптимизация существующего `analysis:bottleneck-detect` (PA1). - **Продуктовые метрики** — плагин `product-management`. - **Документ / change-request процесса** — плагин `operations`. -- Этот скил — про вскрытие **as-is**, не про проектирование. +- **Генерик-методология оптимизации процесса** — скил `process-optimization` + плагина `operations`. Этот скил — про code-grounded discovery конкретного + процесса Лидерры (вскрытие as-is), не про общую методологию и не про + проектирование to-be. ``` - [ ] **Step 2: Create the discovery reference** @@ -483,6 +490,7 @@ git commit -m "feat(c10): add self-authored process-analysis skill (discovery/bo ## Task 5: Bootstrap the C10 home — `docs/process/` + worked example + ADR-008 **Files:** + - Create: `docs/process/README.md`, `docs/process/examples/deal-lifecycle-process.md` - Create: `docs/adr/ADR-008-business-process-tooling.md` - Modify (conditional): `cspell-words.txt` @@ -505,7 +513,7 @@ uses to model, automate and analyse business processes. | Tool | Role | Status | |---|---|---| -| **operations plugin** | Document a process (`/ops:process-doc`), change management (`/ops:change-request`), capacity planning (`/ops:capacity-plan`), vendor review (`/ops:vendor-review`). | installed — `operations@knowledge-work-plugins` | +| **operations plugin** | 9 skills — `process-doc`, `process-optimization`, `change-request`, `capacity-plan`, `compliance-tracking`, `risk-assessment`, `runbook`, `status-report`, `vendor-review`. | installed — `operations@knowledge-work-plugins` v1.2.0 | | **process-modeling skill** | Model a to-be process — BPMN 2.0, process maps, customer-journey / value-stream, RACI, state-machines. | installed — self-authored skill | | **process-analysis skill** | Analyse an as-is process — discovery from code/audit-logs, bottlenecks, traceability, BP metrics. | installed — self-authored skill | | **n8n-mcp** | Workflow-automation engine. | **deferred** — see below | @@ -513,16 +521,21 @@ uses to model, automate and analyse business processes. ## Boundaries (which tool for which job) -- **Documenting a process / a change to it / capacity** → the **operations** plugin. -- **Designing a to-be process model** (BPMN, swimlane, state-machine) → the - **process-modeling** skill — which delegates *rendering* to the **mermaid** skill - (operations and process-modeling emit Mermaid source; mermaid renders — OPS1/BPMN1). -- **Reverse-engineering / optimising an as-is process** → the **process-analysis** - skill. -- `/ops:capacity-plan` plans **cross-functional business-process** capacity; - dev-team sprint capacity is product-management `/sprint-planning` (OPS2). -- `/ops:change-request` = a **business-process** change; an architecture decision - is an ADR (adr-kit); a normative-doc edit is claude-md-management (OPS3). +- **Documenting a process / a change to it / capacity / generic optimization** → + the **operations** plugin skills (`process-doc`, `change-request`, + `capacity-plan`, `process-optimization`, …). +- **Designing a to-be process model** (formal BPMN 2.0, swimlane, state-machine) → + the **process-modeling** skill — which delegates *rendering* to the **mermaid** + skill (operations and process-modeling emit Mermaid source; mermaid renders — + OPS1/BPMN1). +- **Reverse-engineering an as-is process from the Лидерра codebase** → the + **process-analysis** skill (discovery from `routes/`, `app/Jobs`, `audit_*` + tables — what the generic `process-optimization` skill cannot do; OPS5). +- The operations `capacity-plan` skill plans **cross-functional business-process** + capacity; dev-team sprint capacity is product-management `/sprint-planning` (OPS2). +- The operations `change-request` skill = a **business-process** change; an + architecture decision is an ADR (adr-kit); a normative-doc edit is + claude-md-management (OPS3). - process-analysis finds **business-process** bottlenecks; **code/runtime** performance is `perf-analyzer` / `analysis:bottleneck-detect` (PA1). @@ -649,7 +662,7 @@ git commit -m "feat(c10): bootstrap docs/process — README + worked example + A - [ ] **Step 1: Smoke-test the operations plugin** -Run `/ops:process-doc` (or the closest available `/ops:*` command from Task 2 Step 3) on a trivial process (e.g. "лид поступил → менеджер принял → создана сделка"). Expected: the command runs and produces a process document / flowchart. Functional smoke — no file output required. Record which `/ops:*` commands exist. +Invoke the operations `process-doc` skill on a trivial process (e.g. "лид поступил → менеджер принял → создана сделка"). Expected: the skill loads and produces a process document / flowchart. Functional smoke — no file output required. The 9 operations skills were already enumerated on install (Task 2): `process-doc`, `process-optimization`, `change-request`, `capacity-plan`, `compliance-tracking`, `risk-assessment`, `runbook`, `status-report`, `vendor-review`. - [ ] **Step 2: Smoke-test the process-modeling skill** @@ -679,7 +692,7 @@ Read for exact insertion points and the **current** counter: `docs/Tooling_v8_3. Edit `docs/Tooling_v8_3.md`: add `§4.26`–`§4.29` (next four after the last subsection), category **business-process** (off-phase). Per tool: -- **operations** (`#51`) — marketplace plugin `operations@knowledge-work-plugins` (`anthropics/knowledge-work-plugins`), Anthropic Verified; commands `/ops:process-doc` / `/ops:change-request` / `/ops:capacity-plan` / `/ops:vendor-review` (the list confirmed in Task 2 Step 3); boundaries OPS1/OPS2/OPS3; no CC hooks. +- **operations** (`#51`) — marketplace plugin `operations@knowledge-work-plugins` v1.2.0 (`anthropics/knowledge-work-plugins`), Anthropic Verified; ships **9 skills** (`process-doc`, `process-optimization`, `change-request`, `capacity-plan`, `compliance-tracking`, `risk-assessment`, `runbook`, `status-report`, `vendor-review`); **0 lifecycle hooks**; boundaries OPS1/OPS2/OPS3/OPS5. - **process-modeling** (`#52`) — self-authored standalone skill in `.claude/skills/process-modeling/`; BPMN 2.0 + process maps; renders via mermaid (BPMN1); linted, not lint-ignored (LINT1). - **process-analysis** (`#53`) — self-authored standalone skill in `.claude/skills/process-analysis/`; as-is discovery / bottleneck / traceability / metrics; boundary vs perf-analyzer (PA1). - **n8n-mcp** (`#54`) — `czlonkowski/n8n-mcp`, MIT, **pending — NOT installed**; deferred severable task gated on an ADR adopting n8n (N8N1); the Sentry #34 / Jupyter #50 "pending" precedent.