@jakkrichm/create-nexus-devflow 2.0.0 → 2.0.2

package/template/.claude/skills/9arm-skills/scrutinize/SKILL.md

@@ -1,72 +1,72 @@
----
-name: 9arm-skills/scrutinize
-description: Use for plan, PR, and code review. Starts by challenging whether the change should exist, then checks smaller alternatives and traces actual code paths.
-source_pack: 9arm-skills
-credit: thananon/9arm-skills
-upstream: https://github.com/thananon/9arm-skills
-adapted_for: Antigravity IDE / Nexus-DevFlow
----
-
-# Scrutinize
-
-## Purpose
-
-Review with an outsider perspective before accepting a plan, PR, or code change. The first question is not "is this code correct?" but "should this change exist in this form?"
-
-This skill is a credited adaptation of `9arm-skills/scrutinize` for Nexus-DevFlow.
-
-## Review Sequence
-
-### 1. Intent Check
-
-Clarify:
-
-- what problem the change is solving
-- what user or system behavior should change
-- what is explicitly out of scope
-- whether the change matches the task, spec, or PR description
-
-### 2. Smaller / Safer Alternative
-
-Before reviewing details, ask:
-
-- can this be solved with less code?
-- can this reuse an existing pattern?
-- can the risk be split into smaller changes?
-- is any dependency, abstraction, migration, or API surface unnecessary?
-
-### 3. Trace The Actual Path
-
-Do not review only the diff. Follow the end-to-end path that will execute after the change:
-
-- entrypoint
-- existing callers
-- changed module
-- downstream effects
-- persistence or external boundaries
-- error path and fallback behavior
-
-### 4. Findings First
-
-Report actionable findings first, ordered by severity. Avoid style-only comments unless style hides real risk.
-
-## Nexus-DevFlow Output
-
-When used by `PR-Review`, preserve `.agent/resources/schemas/pr_review.template.md` and add scrutiny context without replacing the required sections.
-
-Recommended extra sections:
-
-```md
-## Source Discipline
-- Source pack: 9arm-skills
-- Applied skill: scrutinize
-- Credit: thananon/9arm-skills
-- Adapted for: Antigravity IDE / Nexus-DevFlow
-
-## Intent Check
-## Smaller Alternative Considered
-## Actual Path Trace
-```
-
-For `Agent code-reviewer`, save substantial reports under `devflow/reports/` unless the invoking workflow has a more specific destination.
-
+---
+name: 9arm-skills/scrutinize
+description: Use for plan, PR, and code review. Starts by challenging whether the change should exist, then checks smaller alternatives and traces actual code paths.
+source_pack: 9arm-skills
+credit: thananon/9arm-skills
+upstream: https://github.com/thananon/9arm-skills
+adapted_for: Antigravity IDE / Nexus-DevFlow
+---
+
+# Scrutinize
+
+## Purpose
+
+Review with an outsider perspective before accepting a plan, PR, or code change. The first question is not "is this code correct?" but "should this change exist in this form?"
+
+This skill is a credited adaptation of `9arm-skills/scrutinize` for Nexus-DevFlow.
+
+## Review Sequence
+
+### 1. Intent Check
+
+Clarify:
+
+- what problem the change is solving
+- what user or system behavior should change
+- what is explicitly out of scope
+- whether the change matches the task, spec, or PR description
+
+### 2. Smaller / Safer Alternative
+
+Before reviewing details, ask:
+
+- can this be solved with less code?
+- can this reuse an existing pattern?
+- can the risk be split into smaller changes?
+- is any dependency, abstraction, migration, or API surface unnecessary?
+
+### 3. Trace The Actual Path
+
+Do not review only the diff. Follow the end-to-end path that will execute after the change:
+
+- entrypoint
+- existing callers
+- changed module
+- downstream effects
+- persistence or external boundaries
+- error path and fallback behavior
+
+### 4. Findings First
+
+Report actionable findings first, ordered by severity. Avoid style-only comments unless style hides real risk.
+
+## Nexus-DevFlow Output
+
+When used by `PR-Review`, preserve `.agent/resources/schemas/pr_review.template.md` and add scrutiny context without replacing the required sections.
+
+Recommended extra sections:
+
+```md
+## Source Discipline
+- Source pack: 9arm-skills
+- Applied skill: scrutinize
+- Credit: thananon/9arm-skills
+- Adapted for: Antigravity IDE / Nexus-DevFlow
+
+## Intent Check
+## Smaller Alternative Considered
+## Actual Path Trace
+```
+
+For `Agent code-reviewer`, save substantial reports under `devflow/reports/` unless the invoking workflow has a more specific destination.
+

package/template/.claude/skills/agent/SKILL.md

@@ -1,165 +1,165 @@
 ---name: agent
-
-description: Invoke Agent - Run a specialist persona on a file, folder, stage artifact, or codebase concern.
----
-# Phase 90: Invoke Agent
-
-Run a specialist agent persona for targeted expertise while preserving the DevFlow 2.0 stage-first contract.
-
-Primary behavior now lives in the `specialist-agent-routing` skill. Keep this workflow as the compatibility wrapper and advanced user-facing agent entry point.
-
-## Usage
-
-```text
-Agent {AGENT_NAME} {TARGET}
-```
-
-Examples:
-
-```text
-Agent requirements-engineer devflow/runs/007-auth-refactor/20-spec.md
-Agent codebase-explorer src/services/
-Agent code-reviewer devflow/runs/007-auth-refactor
-```
-
-## Available Agents
-
-Agent definitions live in `../agents/`.
-
-Planning and requirements:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `prp-core-planner` | Create implementation plans from specs and codebase patterns. |
-| `requirements-engineer` | Refine `10-define.md` and `20-spec.md` before planning. |
-| `prp-core-prd-architect` | Draft PRD documents from early ideas. |
-| `orchestrator` | Coordinate complex work and recommend next specialist steps. |
-| `prp-core-boss` | Route `/goal` requests, decompose work, enforce turn budgets, and validate worker output. |
-
-Research and exploration:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `codebase-explorer` | Find files and patterns; analyze architecture, data flow, and module relationships. |
-| `web-researcher` | Research external APIs, docs, and best practices. |
-
-Implementation:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `prp-core-coder` | Implement planned subtasks. |
-| `prp-core-worker` | Execute focused Boss-assigned `/goal` subtasks and report validation status. |
-| `backend-specialist` | Backend, APIs, database, server logic. |
-| `frontend-specialist` | UI/UX and frontend implementation. |
-| `database-architect` | Schema, indexing, and query design. |
-
-Quality and debugging:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `test-engineer` | Unit and integration tests. |
-| `prp-core-debugger` | Root cause analysis. |
-| `code-reviewer` | Code review, risk, regressions. |
-| `security-auditor` | Security and logic risk. |
-| `performance-engineer` | Performance bottlenecks. |
-
-Git and docs:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `prp-core-git-committer` | Commit preparation and message quality. |
-| `prp-core-git-pr-maker` | Pull request summary and preparation. |
-| `documentation-maintainer` | Documentation impact analysis. |
-
-Maintenance and infrastructure:
-
-| Agent | Purpose |
-| :--- | :--- |
-| `ob-loop-engineer` | Overnight infrastructure maintenance, code debt pruning, and variant branch testing loops. |
-
-
-## Reusable Skills
-
-The following former narrow agents are now skills. Invoke them through the responsible agent or workflow instead of `Agent`:
-
-| Skill | Apply Through |
-| :--- | :--- |
-| `code-simplification` | `prp-core-coder`, `code-reviewer`, or `Simplify` |
-| `type-design` | `backend-specialist`, `frontend-specialist`, `database-architect`, or `code-reviewer` |
-| `silent-failure-audit` | `code-reviewer`, `test-engineer`, `backend-specialist`, or `security-auditor` |
-
-## Artifact Contract
-
-Specialist agents must use the DevFlow 2.0 stage-first contract:
-
-- Prefer `00-discover.md`, `10-define.md`, `20-spec.md`, `30-plan.md`, `40-implement.md`, `50-verify.md`, `70-release.md`, and `60-report.md` as the primary artifacts.
-- Do not silently rewrite legacy JSON artifacts.
-- If migration work still requires JSON, prefer CLI-backed minimal updates instead of full-file rewrites.
-- If the specialist is only reviewing, recommend exact commands instead of mutating artifacts.
-- Keep reports readable without requiring a dashboard or JSON viewer.
-
-## 9arm-Skills Discipline Routing
-
-When the target task matches one of these review or communication modes, apply the credited local pack under `.agents/skills/9arm-skills/` while preserving the normal specialist output contract:
-
-| Situation | Apply |
-| :--- | :--- |
-| `code-reviewer`, PR review, plan critique, or change risk review | `9arm-skills/scrutinize` |
-| debug or RCA specialist work | `9arm-skills/debug-mantra` |
-| completed bug or regression lesson extraction | `9arm-skills/post-mortem` |
-| stakeholder-readable summary, PR body, release note, standup, or status update | `9arm-skills/management-talk` |
-
-Reports that use the pack should include a short `Source Discipline` note with credit to `thananon/9arm-skills`.
-
-## Process
-
-1. Load the persona from `../agents/{AGENT_NAME}.md`.
-2. Read the target file, folder, stage directory, or recent changes.
-3. Apply the persona's specialist logic.
-4. If artifact changes are needed, prefer stage markdown updates first.
-5. Record lightweight context usage in the report when the agent loads notable context.
-6. Produce either a concise report or a focused set of changes.
-
-## Output
-
-If a report is substantial, save it to:
-
-```text
-devflow/reports/{date}-{agent-name}-{timestamp}.md
-```
-
-Before saving a substantial agent report, inspect `.agent/resources/schemas/agent_report.template.md` and use its required headings unless the target workflow provides a more specific template.
-
-Before reporting completion, run:
-
-Review `{report_path}` against `agent_report.template.md`, keep the required headings, and remove placeholder text before completion.
-
-Replace any placeholder or template text with concrete files reviewed, findings, recommendations, and validation needs.
-
-For substantial reports, add a short `Context Usage Notes` subsection under the context reviewed area with files or artifacts loaded, any known token counts, and one practical optimization note for future runs.
-
-Short reports can be returned directly in chat.
-
-## Relationship To DevFlow 2.0
-
-- Classification: Companion command
-- Mainline status: Specialist invocation command, not a numbered stage
-- Typical entry points: any stage that needs focused specialist judgment
-- Typical handoff targets: return to the owning workflow such as `/10-Define`, `/20-Spec`, `/30-Plan`, `/40-Implement`, `/50-Verify`, `/60-Report`, `/70-Release`
-
-## Sources
-
-- `AGENTS.md`
-- `.agent/agents/README.md`
-- `.agent/resources/schemas/agent_report.template.md`
-- Related commands: `Research`, `PR-Review`, `Debug`, `Help`, all mainline stages
-
-## Next Workflow Recommendation
-
-- **Primary**: the workflow that owns the target's lifecycle, such as `/10-Define`, `/20-Spec`, `/30-Plan`, `/40-Implement`, `/50-Verify`, `/60-Report`, `/70-Release`, or `PR-Review`.
-- **Why**: `Agent` supplies specialist judgment; the owning workflow should execute the next lifecycle step.
-- **Alternatives**:
-  - `Research` when the specialist work uncovered unresolved external uncertainty
-  - `Wiki` when the specialist report contains reusable project knowledge
-  - `Help {target}` when the correct next route is unclear
-
+
+description: Invoke Agent - Run a specialist persona on a file, folder, stage artifact, or codebase concern.
+---
+# Phase 90: Invoke Agent
+
+Run a specialist agent persona for targeted expertise while preserving the DevFlow 2.0 stage-first contract.
+
+Primary behavior now lives in the `specialist-agent-routing` skill. Keep this workflow as the compatibility wrapper and advanced user-facing agent entry point.
+
+## Usage
+
+```text
+Agent {AGENT_NAME} {TARGET}
+```
+
+Examples:
+
+```text
+Agent requirements-engineer devflow/runs/007-auth-refactor/20-spec.md
+Agent codebase-explorer src/services/
+Agent code-reviewer devflow/runs/007-auth-refactor
+```
+
+## Available Agents
+
+Agent definitions live in `../agents/`.
+
+Planning and requirements:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `prp-core-planner` | Create implementation plans from specs and codebase patterns. |
+| `requirements-engineer` | Refine `10-define.md` and `20-spec.md` before planning. |
+| `prp-core-prd-architect` | Draft PRD documents from early ideas. |
+| `orchestrator` | Coordinate complex work and recommend next specialist steps. |
+| `prp-core-boss` | Route `/goal` requests, decompose work, enforce turn budgets, and validate worker output. |
+
+Research and exploration:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `codebase-explorer` | Find files and patterns; analyze architecture, data flow, and module relationships. |
+| `web-researcher` | Research external APIs, docs, and best practices. |
+
+Implementation:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `prp-core-coder` | Implement planned subtasks. |
+| `prp-core-worker` | Execute focused Boss-assigned `/goal` subtasks and report validation status. |
+| `backend-specialist` | Backend, APIs, database, server logic. |
+| `frontend-specialist` | UI/UX and frontend implementation. |
+| `database-architect` | Schema, indexing, and query design. |
+
+Quality and debugging:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `test-engineer` | Unit and integration tests. |
+| `prp-core-debugger` | Root cause analysis. |
+| `code-reviewer` | Code review, risk, regressions. |
+| `security-auditor` | Security and logic risk. |
+| `performance-engineer` | Performance bottlenecks. |
+
+Git and docs:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `prp-core-git-committer` | Commit preparation and message quality. |
+| `prp-core-git-pr-maker` | Pull request summary and preparation. |
+| `documentation-maintainer` | Documentation impact analysis. |
+
+Maintenance and infrastructure:
+
+| Agent | Purpose |
+| :--- | :--- |
+| `ob-loop-engineer` | Overnight infrastructure maintenance, code debt pruning, and variant branch testing loops. |
+
+
+## Reusable Skills
+
+The following former narrow agents are now skills. Invoke them through the responsible agent or workflow instead of `Agent`:
+
+| Skill | Apply Through |
+| :--- | :--- |
+| `code-simplification` | `prp-core-coder`, `code-reviewer`, or `Simplify` |
+| `type-design` | `backend-specialist`, `frontend-specialist`, `database-architect`, or `code-reviewer` |
+| `silent-failure-audit` | `code-reviewer`, `test-engineer`, `backend-specialist`, or `security-auditor` |
+
+## Artifact Contract
+
+Specialist agents must use the DevFlow 2.0 stage-first contract:
+
+- Prefer `00-discover.md`, `10-define.md`, `20-spec.md`, `30-plan.md`, `40-implement.md`, `50-verify.md`, `70-release.md`, and `60-report.md` as the primary artifacts.
+- Do not silently rewrite legacy JSON artifacts.
+- If migration work still requires JSON, prefer CLI-backed minimal updates instead of full-file rewrites.
+- If the specialist is only reviewing, recommend exact commands instead of mutating artifacts.
+- Keep reports readable without requiring a dashboard or JSON viewer.
+
+## 9arm-Skills Discipline Routing
+
+When the target task matches one of these review or communication modes, apply the credited local pack under `.agents/skills/9arm-skills/` while preserving the normal specialist output contract:
+
+| Situation | Apply |
+| :--- | :--- |
+| `code-reviewer`, PR review, plan critique, or change risk review | `9arm-skills/scrutinize` |
+| debug or RCA specialist work | `9arm-skills/debug-mantra` |
+| completed bug or regression lesson extraction | `9arm-skills/post-mortem` |
+| stakeholder-readable summary, PR body, release note, standup, or status update | `9arm-skills/management-talk` |
+
+Reports that use the pack should include a short `Source Discipline` note with credit to `thananon/9arm-skills`.
+
+## Process
+
+1. Load the persona from `../agents/{AGENT_NAME}.md`.
+2. Read the target file, folder, stage directory, or recent changes.
+3. Apply the persona's specialist logic.
+4. If artifact changes are needed, prefer stage markdown updates first.
+5. Record lightweight context usage in the report when the agent loads notable context.
+6. Produce either a concise report or a focused set of changes.
+
+## Output
+
+If a report is substantial, save it to:
+
+```text
+devflow/reports/{date}-{agent-name}-{timestamp}.md
+```
+
+Before saving a substantial agent report, inspect `.agent/resources/schemas/agent_report.template.md` and use its required headings unless the target workflow provides a more specific template.
+
+Before reporting completion, run:
+
+Review `{report_path}` against `agent_report.template.md`, keep the required headings, and remove placeholder text before completion.
+
+Replace any placeholder or template text with concrete files reviewed, findings, recommendations, and validation needs.
+
+For substantial reports, add a short `Context Usage Notes` subsection under the context reviewed area with files or artifacts loaded, any known token counts, and one practical optimization note for future runs.
+
+Short reports can be returned directly in chat.
+
+## Relationship To DevFlow 2.0
+
+- Classification: Companion command
+- Mainline status: Specialist invocation command, not a numbered stage
+- Typical entry points: any stage that needs focused specialist judgment
+- Typical handoff targets: return to the owning workflow such as `/10-Define`, `/20-Spec`, `/30-Plan`, `/40-Implement`, `/50-Verify`, `/60-Report`, `/70-Release`
+
+## Sources
+
+- `AGENTS.md`
+- `.agent/agents/README.md`
+- `.agent/resources/schemas/agent_report.template.md`
+- Related commands: `Research`, `PR-Review`, `Debug`, `Help`, all mainline stages
+
+## Next Workflow Recommendation
+
+- **Primary**: the workflow that owns the target's lifecycle, such as `/10-Define`, `/20-Spec`, `/30-Plan`, `/40-Implement`, `/50-Verify`, `/60-Report`, `/70-Release`, or `PR-Review`.
+- **Why**: `Agent` supplies specialist judgment; the owning workflow should execute the next lifecycle step.
+- **Alternatives**:
+  - `Research` when the specialist work uncovered unresolved external uncertainty
+  - `Wiki` when the specialist report contains reusable project knowledge
+  - `Help {target}` when the correct next route is unclear
+