AGENT SKILL

x-create-feature

Create a complete feature (Epic + N Stories + Implementation Map) from a spec file, with a worktree-isolated docs/ branch, consolidated commit, and auto-merged PR into epic/XXXX.

License 许可
MIT
最近更新
2026-10-27
Stars
★ 0
Forks
1

01 前言 Front Matter

name
x-create-feature
model
sonnet
description
Create a complete feature (Epic + N Stories + Implementation Map) from a spec file, with a worktree-isolated docs/ branch, consolidated commit, and auto-merged PR into epic/XXXX.
allowed-tools
Read, Write, Edit, Bash, Grep, Glob, AskUserQuestion, Skill
argument-hint
[SPEC-FILE-PATH] --epic-id <NNNN> [--no-jira] [--dry-run]
context-budget
medium
user-invocable
true
requires-capabilities
[]

02 资源文件 1

03 SKILL.md 正文

Output Policy

  • Language: Portuguese (pt-BR) for all content. English for technical terms (cache, timeout, handler, endpoint) and code identifiers.
  • Tone: Technical, Direct, and Concise.
  • Efficiency: Remove all conversational fillers and greetings to save tokens.

Skill: x-create-feature (Feature Creation Chain Orchestrator)

Purpose

Orchestrate the full creation of a feature from a system specification: an Epic, individual Story files, and an Implementation Map. The entire operation runs inside an isolated worktree (feature-XXXX-<slug>), producing a docs/<epic-id>-<slug> branch with a single consolidated commit, auto-merged into epic/XXXX via PR.

EPIC-0065 hard-cut (Rule 19 §Hard-cut autorizado): This skill supersedes x-epic-decompose (removed). Internal sub-skills x-internal-create-epic, x-internal-map-epic, and x-internal-create-story are now the implementation of the generation phases (not user-invocable).

When to Use

  • Create a complete feature backlog from a spec: epic + stories + implementation map
  • After reviewing the spec PR produced by x-ideate-feature
  • User wants to turn a specification into immediately implementable work items

Prerequisites

Read before starting:

  • .claude/templates/_TEMPLATE-EPIC.md (RA9 v2: 9 sections)
  • .claude/templates/_TEMPLATE-STORY.md (RA9 v2: 9 sections)
  • .claude/templates/_TEMPLATE-IMPLEMENTATION-MAP.md
  • .claude/skills/planning-standards-kp/SKILL.md — Mandatory RA9 9-section model (source of truth), granularity per level (Epic/Story/Task), Packages Hexagonal catalog format, Decision Rationale micro-template
  • references/decomposition-guide.md (bundled with this skill)

If any template is missing, stop and tell the user.

Workflow Overview

P1.5 WORKTREE      -> x-internal-precheck-worktree + x-manage-worktrees create feature-XXXX-<slug>
P2.  BRANCH        -> x-internal-ensure-epic-branch --epic-id XXXX (RULE-001)
1.   ANALYSIS      -> Read spec, identify rules, stories, dependencies, phases (inline)
1.5  JIRA          -> Determine Jira integration mode (conditional, inline)
2.   EPIC          -> x-internal-create-epic instructions (Phase 2 below)
3.   STORIES       -> x-internal-create-story instructions (Phase 3 below)
4.   MAP           -> x-internal-map-epic instructions (Phase 4 below)
4.5  JIRA LINKS    -> Create Jira dependency links (conditional, inline)
P4.  COMMIT        -> x-commit-planning consolidated commit on docs/<epic-id>-<slug>
P5.  PUSH          -> git push docs/<epic-id>-<slug> to origin
P6.  PR            -> x-create-pr --target-branch epic/XXXX --auto-merge merge --label docs
P7.  CI-WATCH      -> x-watch-pr-ci MANDATORY (Rule 45)
5.   REPORT        -> Save all files, validate quality, report summary

Note on consolidated commit

x-create-feature uses a single consolidated commit in Phase P4 covering the full batch (epic + N stories + implementation map). Sub-skill phases write artifacts but do NOT commit individually — the orchestrator owns the commit on branch docs/<epic-id>-<slug>, not on epic/XXXX directly.


Phase P1.5 — Worktree Setup (RULE-001 + EPIC-0065 D-R5)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P1_5-Worktree

P1.5.1 — Idempotent re-entrancy check:

Skill(skill: "x-internal-precheck-worktree", model: "haiku", args: "--identifier feature-<epic-id>-<slug> --allow-dirty")
  • Exit 0 (CLEAN): proceed to worktree creation.
  • Exit 15 (WORKTREE_AMBIGUOUS): log warning and proceed with caution.
  • On non-zero for other reasons: abort with WORKTREE_FAILED.

P1.5.2 — Create feature worktree:

Derive slug from the spec file title or filename (kebab-case, max 40 chars, ASCII-only).

Skill(skill: "x-manage-worktrees", model: "haiku", args: "create --identifier feature-<epic-id>-<slug> --branch docs/<epic-id>-<slug> --base epic/<epic-id>")
  • Worktree path: .claude/worktrees/feature-<epic-id>-<slug>/
  • Branch: docs/<epic-id>-<slug> (base: epic/XXXX, per D-R6 — docs PRs auto-merge into epic branch)
  • All subsequent file writes happen inside this worktree.
  • Worktree is creator-owned (Rule 14 §Invariants) — removed only after successful PR merge.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P1_5-Worktree ok


Phase P2 — Ensure epic/<ID> Branch (RULE-001)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P2-Epic-Branch

Ensure the canonical epic/<ID> branch exists locally AND on origin before writing any artifact:

Skill(skill: "x-internal-ensure-epic-branch", model: "haiku", args: "--epic-id <XXXX>")

Abort with EPIC_BRANCH_ENSURE_FAILED on any non-zero exit.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P2-Epic-Branch ok


Phase 1 — Analysis

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-1-Analysis

  1. Read the spec file (path from argument or the spec PR produced by x-ideate-feature)
  2. Read references/decomposition-guide.md
  3. Analyze the spec:
    • Identify cross-cutting rules (spanning multiple journeys)
    • Identify stories by layer (foundation → core → extensions → compositions → cross-cutting)
    • Map dependencies between stories
    • Compute phases from the dependency DAG
    • Identify the critical path

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-1-Analysis ok


Phase 1.5 — Jira Integration Decision

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-1_5-Jira

Before generating artifacts, determine Jira integration mode. Logic unchanged from predecessor:

  • If --no-jira: jiraContext = { enabled: false }, skip.
  • If --jira <PROJECT_KEY>: discover cloudId via mcp__atlassian__getAccessibleAtlassianResources, set jiraContext.
  • If no flag: prompt via AskUserQuestion (3 options: all Jira / epic only / markdown only).

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-1_5-Jira ok


Phase 2 — Generate the Epic (x-internal-create-epic)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-2-Epic

Follow the instructions in .claude/skills/x-internal-create-epic/SKILL.md:

  • Determine epic number (scan ai/epics/ for existing epic-XXXX-* folders; use --epic-id override when supplied)
  • Create directory ai/epics/epic-XXXX-<slug>/ inside the worktree
  • Extract rules → RULE-001..N table
  • Build story index with titles and dependencies
  • Define DoR/DoD (must include TDD Compliance + Double-Loop TDD)
  • Generate ai/epics/epic-XXXX-<slug>/epic-XXXX.md following _TEMPLATE-EPIC.md

Jira integration: same logic as predecessor (create Epic issue if jiraContext.enabled).

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-2-Epic ok


Phase 3 — Generate the Stories (x-internal-create-story)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-3-Stories

Follow the instructions in .claude/skills/x-internal-create-story/SKILL.md:

For each story in the Epic index:

  • Dependencies, applicable rules, user story description, technical context
  • Entrega de Valor (Section 3.5) — measurable business value
  • Data contracts (precise field names, types, formats)
  • Mermaid sequence diagrams (real component names)
  • Gherkin acceptance criteria (all 4 mandatory categories, TPP order)
  • Sub-tasks tagged [Dev], [Test], [Doc] — at least one [Test] Smoke/E2E

Generate files as ai/epics/epic-XXXX-<slug>/story-XXXX-YYYY.md following _TEMPLATE-STORY.md.

Jira integration: same cascade logic as predecessor.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-3-Stories ok


Phase 4 — Generate the Implementation Map (x-internal-map-epic)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-4-Map

Follow the instructions in .claude/skills/x-internal-map-epic/SKILL.md:

  • Dependency matrix (all stories, validated for consistency)
  • Phase diagram (ASCII box-drawing)
  • Critical path analysis
  • Mermaid dependency graph (color-coded by phase)
  • Phase summary and detail tables
  • Strategic observations

Generate ai/epics/epic-XXXX-<slug>/IMPLEMENTATION-MAP.md following _TEMPLATE-IMPLEMENTATION-MAP.md.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-4-Map ok


Phase 4.5 — Jira Dependency Linking (conditional)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-4_5-Jira-Links

If jiraContext.enabled == true and jiraContext.cascadeToStories == true: create Jira dependency links between stories via mcp__atlassian__createIssueLink. Best-effort — failures are warned, not fatal.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-4_5-Jira-Links ok


Phase P4 — Consolidated Planning Commit (RULE-007)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P4-Commit

If --dry-run: log "dry-run, skipping commit" and skip.

Collect the full path set (epic + all stories + map) and commit via x-commit-planning:

Skill(skill: "x-commit-planning", model: "haiku",
      args: "--scope chore --epic-id <XXXX> --paths ai/epics/epic-<XXXX>-<slug>/epic-<XXXX>.md,ai/epics/epic-<XXXX>-<slug>/story-*.md,ai/epics/epic-<XXXX>-<slug>/IMPLEMENTATION-MAP.md --subject \"full decomposition (epic + <N> stories + map)\" --branch docs/<epic-id>-<slug>")

On COMMIT_FAILED (exit 4): abort. On noOp=true: silent no-op.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P4-Commit ok


Phase P5 — Push docs Branch

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P5-Push

If --dry-run: skip.

Push the docs branch to origin:

git push -u origin docs/<epic-id>-<slug>

On push failure: log WARNING and continue. Local commit preserved.

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P5-Push ok


Phase P5.5 — BACKLOG-RENDER (EPIC-0066, story-0066-0006)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P55-Render

MANDATORY TOOL CALL — Rule 24 §Camada-1 / Rule 13 Pattern 1 (INLINE-SKILL):

Render the backlog PR body from the committed epic.md + IMPLEMENTATION-MAP.md + stories. This produces a body Markdown that includes the <!-- template-version: 1.0 --> marker required by audit-pr-template.sh (story-0066-0007).

TMP_BODY_PATH=$(mktemp -t feature-pr-body-XXXXXX.md)
chmod 600 "$TMP_BODY_PATH"
trap 'rm -f "$TMP_BODY_PATH"' EXIT
echo "INFO: Invoking x-internal-render-pr-body --kind=backlog --epic-id=$EPIC_ID"

Invoke the x-internal-render-pr-body skill via the Skill tool:

Skill(skill: "x-internal-render-pr-body", model: "haiku", args: "--kind=backlog --epic-id=epic-<EPIC_ID> --out=$TMP_BODY_PATH")
RENDER_EXIT=$?
if [[ $RENDER_EXIT -eq 0 ]]; then
  PR_BODY_FILE="$TMP_BODY_PATH"   # rendered body with template-version marker
else
  echo "WARN [render-fallback]: x-internal-render-pr-body falhou (exit $RENDER_EXIT). Usando body manual." >&2
  echo "     audit-pr-template.sh reportará PR_TEMPLATE_VIOLATION para este PR." >&2
  PR_BODY_FILE=""   # empty → fallback path in Phase P6
fi

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P55-Render ok


Phase P6 — PR Creation (docs → epic/XXXX)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P6-PR

If --dry-run: skip.

Create the docs PR using the body rendered in Phase P5.5. Auto-merge is authorized for docs/<epic-id>-* → epic/XXXX per Rule 21 §Anti-Patterns exception. The flow uses gh pr create --body-file (not the x-create-pr skill) because x-create-pr is task-scoped and renders --kind=implementation; backlog PRs need --kind=backlog (story-0066-0006).

PR_TITLE="chore(epic-${EPIC_ID}): full decomposition (epic + ${STORY_COUNT} stories + map)"

if [[ -n "$PR_BODY_FILE" ]]; then
  # Happy path — use rendered body with template-version marker
  PR_URL=$(gh pr create \
    --base "epic/${EPIC_ID}" \
    --head "docs/${EPIC_ID}-${SLUG}" \
    --title "$PR_TITLE" \
    --body-file "$PR_BODY_FILE" \
    --label docs --label "epic-${EPIC_ID}")
else
  # Fallback path — see ## Recovery (manual body without template-version marker)
  PR_URL=$(gh pr create \
    --base "epic/${EPIC_ID}" \
    --head "docs/${EPIC_ID}-${SLUG}" \
    --title "$PR_TITLE" \
    --body "$(cat <<EOF
chore(epic-${EPIC_ID}): full decomposition (epic + ${STORY_COUNT} stories + map)

(rendered via fallback path — body lacks template-version marker; audit will block merge)
EOF
)" \
    --label docs --label "epic-${EPIC_ID}")
fi

# Enable auto-merge (Rule 21 §Anti-Patterns docs/* exception)
PR_NUMBER=$(echo "$PR_URL" | grep -oE '[0-9]+$')
gh pr merge "$PR_NUMBER" --merge --auto

# Capture {prUrl, prNumber} for Phase P7

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P6-PR ok


Phase P7 — CI-Watch (Rule 45 — MANDATORY)

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-P7-CIWatch

MANDATORY TOOL CALL — NON-NEGOTIABLE (Rule 45): Invoke x-watch-pr-ci to wait for CI + merge status:

Skill(skill: "x-watch-pr-ci", model: "haiku", args: "--pr-number <prNumber>")

Persist .claude/state/pr-watch-{prNumber}.json (Rule 45 evidence).

Dispatch on exit code:

  • SUCCESS (0): CI green + PR merged → proceed to Phase 5 (report)
  • CI_PENDING_PROCEED (10): CI green, no Copilot review → proceed with WARNING
  • CI_FAILED (20): block; log CI_FAILED and abort
  • TIMEOUT (30): surface to operator; abort
  • PR_ALREADY_MERGED (40): idempotent → proceed
  • NO_CI_CONFIGURED (50): skip CI gate; proceed
  • PR_CLOSED (60): abort
  • PR_NOT_FOUND (70): abort

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-P7-CIWatch ok


Phase 5 — Save and Report

<!-- TELEMETRY: phase.start --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh start x-create-feature Phase-5-Report

All files saved inside ai/epics/epic-XXXX-<slug>/ (v4 layout — D-R1). Quality validation:

  • [ ] Every story has Entrega de Valor (measurable, not technical)
  • [ ] Every story has at least one [Test] Smoke/E2E sub-task
  • [ ] Every story DoD includes automated test requirement
  • [ ] Dependencies are symmetric
  • [ ] No circular dependencies
  • [ ] Phase computation is correct

Report summary:

  • Total rules extracted, stories generated, phases computed, critical path, max parallelism
  • PR: #<prNumber> — <prUrl> — Status: MERGED (or AWAITING CI)
  • Jira summary (if applicable)

Print:

====================================================
 x-create-feature — Feature ready
====================================================
 Epic:    ai/epics/epic-XXXX-<slug>/epic-XXXX.md
 Stories: N files under ai/epics/epic-XXXX-<slug>/
 Map:     ai/epics/epic-XXXX-<slug>/IMPLEMENTATION-MAP.md
 Branch:  docs/<epic-id>-<slug>
 PR:      #<prNumber> — <prUrl> — MERGED
----------------------------------------------------
 Next step: /x-implement-epic XXXX
====================================================

<!-- TELEMETRY: phase.end --> Bash command: $CLAUDE_PROJECT_DIR/.claude/hooks/telemetry-phase.sh end x-create-feature Phase-5-Report ok


Recovery (Render Skill Fallback — EPIC-0066, story-0066-0006)

When x-internal-render-pr-body returns exit ≠ 0 in Phase P5.5, this skill falls back to a minimal manual body in Phase P6 to ensure PR creation never aborts (RULE-004 fail-open). The fallback body lacks the <!-- template-version: 1.0 --> marker — audit-pr-template.sh (story-0066-0007) will block the merge with PR_TEMPLATE_VIOLATION. This is intentional: the preventive render path produces the marker; the detective audit gate enforces it.

The WARN message includes RENDER_EXIT (the render skill exit code) so the operator can diagnose the cause: typically OPERATIONAL_ERROR (exit 2) when _TEMPLATE-PR-BACKLOG.md is missing from .claude/templates/ (rerun mvn process-resources).

Error Handling

Scenario Action
Spec file not found or empty Abort: Spec file not found or is empty.
Template file missing Stop and tell the user which template is missing
x-internal-precheck-worktree exit 15 Log WARNING, proceed with caution
x-manage-worktrees fails Abort with WORKTREE_FAILED
x-internal-ensure-epic-branch fails Abort with EPIC_BRANCH_ENSURE_FAILED
x-commit-planning exit 4 Abort with COMMIT_FAILED
x-commit-planning noOp=true Silent no-op, continue to P5
x-create-pr fails Abort with PR_CREATE_FAILED
x-watch-pr-ci exit 20 (CI_FAILED) Abort; preserve worktree for diagnosis
--dry-run set Phases P4, P5, P6, P7 are no-ops
Circular dependency detected Abort with cycle description
Story fails quality validation Fix before saving — do not skip validation

Integration Notes

Skill Relationship Notes
x-internal-precheck-worktree delegates (P1.5.1) Idempotent re-entrancy check
x-manage-worktrees delegates (P1.5.2) Creates feature-XXXX-<slug> worktree
x-internal-ensure-epic-branch delegates (P2) Ensures epic/XXXX on local + origin
x-internal-create-epic follows instructions (Ph 2) Generates Epic file
x-internal-create-story follows instructions (Ph 3) Generates Story files
x-internal-map-epic follows instructions (Ph 4) Generates Implementation Map
x-commit-planning delegates (P4) Single consolidated commit on docs branch
x-create-pr delegates (P6) Opens PR with auto-merge (Rule 21 docs/ exception)
x-watch-pr-ci delegates (P7) MANDATORY CI-watch (Rule 45)
x-ideate-feature preceded by Optional — produces spec PR reviewed by operator
x-implement-epic followed by Implements the generated epic

Backward Compatibility (EPIC-0065)

x-epic-decompose was removed via hard-cut (Rule 19 §Hard-cut autorizado — RULE-004). No alias. Operators update invocations to /x-create-feature. Internal sub-skills (x-internal-create-epic, x-internal-map-epic, x-internal-create-story) are not user-invocable (Rule 22 — visibility: internal).

04 同仓库更多 Skills

review-bench
review-bench/edercnj_ia-dev-environment

x-parallel-eval

Detects and classifies file-collision risks between work units (tasks, stories, or epic phases) before parallel execution. Reads ## File Footprint blocks produced by x-task-plan / x-story-plan, applies the parallelism-heuristics knowledge pack (hard / regen / soft categories + hotspot overrides), and emits a collision matrix + serialization recommendation in Markdown (default) or JSON.

★ 0
review-bench
review-bench/edercnj_ia-dev-environment

parallelism-heuristics

Canonical catalog of file-collision heuristics for parallel task, story, and epic execution. Defines the File Footprint block format, the three Conflict Categories (hard / regen / soft), a hotspot list of high-contention paths, and the degrade-with-warning policy. Consumed by x-task-plan, x-story-plan, x-epic-map, x-parallel-eval, and x-dev-*-implement.

★ 0
review-bench
review-bench/edercnj_ia-dev-environment

x-watch-pr-ci

Polls a PR's CI checks and Copilot review status, blocking until checks complete or timeout. Returns one of 8 stable exit codes (SUCCESS=0, CI_PENDING_PROCEED=10, CI_FAILED=20, TIMEOUT=30, PR_ALREADY_MERGED=40, NO_CI_CONFIGURED=50, PR_CLOSED=60, PR_NOT_FOUND=70). Writes a versioned state-file for session resume.

★ 0
review-bench
review-bench/edercnj_ia-dev-environment

x-validate-docs

Documentation freshness gate: validates 6 dimensions (readme, api-specs, grpc-proto, adr, skill-docs, system-architecture) against code changes in a PR. Stack-aware — only validates targets declared/auto-detected from ProjectConfig.documentation.targets. Returns exit non-zero on staleness; produces structured report.

★ 0
review-bench
review-bench/edercnj_ia-dev-environment

x-update-system-architecture

Incrementally updates docs/architecture/system.md after an epic completes. Appends new entries to the Decision Log (§11) using x-internal-write-report --append (dedup by ## ID: marker) and surgically inserts component/integration changes into sections 1-10 via Edit. Idempotent: re-running with the same epic produces a byte-identical system.md.

★ 0
review-bench
review-bench/edercnj_ia-dev-environment

x-update-architecture

Incrementally updates the service architecture document with changes from architecture plans. Adds new components, integrations, flows, and ADR references without rewriting existing content. Use after implementation to keep architecture documentation current.

★ 0