| name | cc-version-analysis |
| classification | workflow |
| classification-reason | Orchestrates multi-phase research and analysis pipeline independent of model capability |
| deprecation-risk | none |
| effort | high |
| description | CC CLI version upgrade impact analysis โ research changes, analyze bkit impact, generate report.
Triggers: cc-version-analysis, CC upgrade, version analysis, CC ๋ฒ์ ๋ถ์, ๋ฒ์ ์ํฅ.
|
| argument-hint | [from_version] [to_version] |
| user-invocable | true |
| agents | {"research":"bkit:cc-version-researcher","analyze":"bkit:bkit-impact-analyst","report":"bkit:report-generator"} |
| allowed-tools | ["Read","Write","Edit","Glob","Grep","Bash","Task","TaskCreate","TaskUpdate","TaskList","TaskGet","AskUserQuestion","Agent","WebSearch","WebFetch"] |
| imports | ["${PLUGIN_ROOT}/templates/cc-version-analysis.template.md","${PLUGIN_ROOT}/templates/plan-plus.template.md","${PLUGIN_ROOT}/templates/report.template.md"] |
| next-skill | pdca plan |
| pdca-phase | null |
| task-template | [CC-Version-Analysis] CC v{from} โ v{to} |
CC Version Analysis โ Claude Code CLI ๋ฒ์ ์ํฅ ๋ถ์ ์ํฌํ๋ก์ฐ
CC CLI ๋ฒ์ ์
๊ทธ๋ ์ด๋ ์ bkit plugin์ ๋ํ ์ํฅ์ ์ฒด๊ณ์ ์ผ๋ก ์กฐ์ฌ, ๋ถ์ํ๊ณ
๊ฐ์ ๊ธฐํ๋ฅผ ๋์ถํ๋ ์ ๋ฌธ ์ํฌํ๋ก์ฐ ์คํฌ.
Overview
์ด ์คํฌ์ CC CLI์ ์ ๋ฒ์ ์ด ์ถ์๋์์ ๋ ๋ค์์ ์๋ํํฉ๋๋ค:
- Phase 1 (Research): CC ๋ณ๊ฒฝ์ฌํญ ์ฌ์ธต ์กฐ์ฌ
- Phase 2 (Analyze): bkit ์ํคํ
์ฒ ์ํฅ ๋ถ์
- Phase 3 (Brainstorm): Plan Plus ๋ธ๋ ์ธ์คํ ๋ฐ์ผ๋ก ๊ฐ์ ์ ๋์ถ
- Phase 4 (Report): ์ข
ํฉ ์ํฅ ๋ถ์ ๋ณด๊ณ ์ ์์ฑ
Agent Team ๊ตฌ์ฑ:
cc-version-researcher: CC ๋ฒ์ ๋ณ๊ฒฝ์ฌํญ ์ธ๋ถ ์กฐ์ฌ
bkit-impact-analyst: bkit ๋ด๋ถ ์ํคํ
์ฒ ์ํฅ ๋ถ์
report-generator: ์ต์ข
๋ณด๊ณ ์ ์์ฑ
HARD-GATE
Do NOT skip any phase. Each phase produces artifacts that feed into the next.
Do NOT generate the final report without completing Research and Analysis phases.
Do NOT implement any ENH items โ this skill is analysis-only.
All documents MUST be written in Korean (ํ๊ตญ์ด).
NEW (v2.1.16 errata learning โ added 2026-05-20):
Do NOT advance from Phase 1 โ Phase 2 without completing the Phase 1.5
Raw Source Verification Gate. The cc-version-researcher agent's quantitative
claims (bullet count, issue counts, file counts) MUST be cross-checked against
โฅ2 raw sources (GitHub release tag URL + raw CHANGELOG.md URL) before being
adopted into the report. The v2.1.145 cycle leaked an unverified bullet count
of 21 (actual: 20) and an unverified agent count of 36 (actual: 34) because
this gate did not exist. If raw and agent disagree, raw wins.
Invocation
/cc-version-analysis # Auto-detect: installed vs latest
/cc-version-analysis 2.1.78 2.1.85 # Specific version range
/cc-version-analysis to 2.1.85 # From current installed to target
Process Flow
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ CC Version Analysis โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ โ
โ Phase 0: Setup & Version Detection โ
โ โโโ Detect installed CC version (claude --version) โ
โ โโโ Determine target version (args or latest) โ
โ โโโ Create Task tracking structure โ
โ โโโ Load previous analysis from memory โ
โ โ
โ Phase 1: Research (cc-version-researcher agent) โ
โ โโโ Official docs (code.claude.com) โ
โ โโโ GitHub (anthropics/claude-code) โ
โ โ โโโ Releases & changelog โ
โ โ โโโ Issues (open & recently closed) โ
โ โ โโโ PRs (merged in version range) โ
โ โ โโโ Commits (significant changes) โ
โ โโโ npm registry (@anthropic-ai/claude-code) โ
โ โโโ Technical blogs & community โ
โ โโโ Output: CC Change Report (structured) โ
โ โ
โ Phase 1.5: Raw Source Verification Gate (MANDATORY) โ
โ โโโ WebFetch https://github.com/anthropics/ โ
โ โ claude-code/releases/tag/v{to} โ
โ โโโ WebFetch https://raw.githubusercontent.com/ โ
โ โ anthropics/claude-code/main/CHANGELOG.md โ
โ โโโ Cross-check agent's bullet count vs raw count โ
โ โโโ If mismatch โ raw wins, record errata โ
โ โโโ Spot-check โฅ3 random Fixed bullets verbatim โ
โ โโโ Block Phase 2 if any mismatch unresolved โ
โ โ
โ Phase 2: Analyze (bkit-impact-analyst agent) โ
โ โโโ Map CC changes โ bkit components โ
โ โโโ Identify ENH opportunities โ
โ โโโ File impact matrix โ
โ โโโ Philosophy compliance check โ
โ โโโ Test impact assessment โ
โ โโโ Output: bkit Impact Analysis (structured) โ
โ โ
โ Phase 3: Brainstorm (Plan Plus methodology) โ
โ โโโ Intent discovery (ํต์ฌ ๋ชฉํ/๋ฆฌ์คํฌ/๊ธฐํ) โ
โ โโโ Alternative exploration โ
โ โโโ YAGNI review (๊ฐ ENH ํ์์ฑ ๊ฒ์ฆ) โ
โ โโโ Priority assignment (P0~P3) โ
โ โโโ Output: Prioritized ENH roadmap โ
โ โ
โ Phase 4: Report Generation โ
โ โโโ Merge all phase outputs โ
โ โโโ Generate from template โ
โ โ (cc-version-analysis.template.md) โ
โ โโโ Save to docs/04-report/features/ โ
โ โโโ Update MEMORY.md (version history) โ
โ โโโ Output: Final Impact Report (Korean) โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Phase Details
Phase 0: Setup & Version Detection
1. Detect installed CC version:
$ claude --version
2. Determine target version:
- If args provided: use specified versions
- If no args: search for latest available version
3. Create Task structure:
TaskCreate: "[CC-Version-Analysis] CC v{from} โ v{to}"
โโโ Task: "Phase 1: CC ๋ณ๊ฒฝ์ฌํญ ์กฐ์ฌ"
โโโ Task: "Phase 2: bkit ์ํฅ ๋ถ์"
โโโ Task: "Phase 3: Plan Plus ๋ธ๋ ์ธ์คํ ๋ฐ"
โโโ Task: "Phase 4: ๋ณด๊ณ ์ ์์ฑ"
4. Load previous analysis context:
- Read memory/cc_version_history_*.md
- Read last ENH number from MEMORY.md
- Read existing PDCA status from .bkit/state/pdca-status.json
Phase 1: Research (Agent: cc-version-researcher)
Input: from_version, to_version
Output: Structured CC Change Report
Launch the cc-version-researcher agent with:
Research CC CLI changes from v{from} to v{to}.
Sources: official docs, GitHub (issues/PRs/releases), npm, blogs.
Categorize by: Breaking/Feature/Fix/Performance/SystemPrompt/Hook/Config.
Rate impact: HIGH/MEDIUM/LOW.
Flag bkit-relevant changes.
Output structured markdown tables.
Parallel research tasks (when using Agent Team):
- Task 1: GitHub releases + changelog
- Task 2: GitHub issues (open + recently closed)
- Task 3: Official docs changes
- Task 4: System prompt diff analysis
Phase 1.5: Raw Source Verification Gate (MANDATORY โ added 2026-05-20)
Why: cc-version-researcher's output may paraphrase, summarize, or miscount.
The v2.1.145 cycle revealed that a single WebFetch (via the model-processed
release tag page) under-counted by 1 bullet (reported 6 Added, actual 7).
A second fetch against raw CHANGELOG.md is required to catch this.
Protocol (main session, not delegated):
- Fetch raw GitHub release tag page:
WebFetch https://github.com/anthropics/claude-code/releases/tag/v{to_version}
- Fetch raw CHANGELOG.md (authoritative):
WebFetch https://raw.githubusercontent.com/anthropics/claude-code/main/CHANGELOG.md
prompt: "Show the FULL v{to_version} section verbatim. List every bullet
exactly as written, in order, under the original headings (Added/Fixed/
Improved/Breaking/etc). Do NOT summarize."
- Compare to cc-version-researcher's reported counts:
- Bullet count (total / per-heading)
- Breaking count
- HIGH/MEDIUM/LOW split
- Key item presence (security fixes, OTEL, hook input changes)
- If raw and agent disagree โ raw wins. Record discrepancy in report
under ยง3.0 "Verification Notes" with both numbers.
- Spot-check โฅ3 random Fixed bullets โ verify verbatim text in agent report
matches raw.
- Block Phase 2 progression if any mismatch is unresolved.
Output: ยง3.0 Verification table appended to Phase 1 output:
| Field | Agent reported | Raw verified | Source URL | Verdict |
|---|
| Added | N | M | raw CHANGELOG | match / errata |
| Fixed | N | M | raw CHANGELOG | match / errata |
| Improved | N | M | raw CHANGELOG | match / errata |
| Breaking | N | M | raw CHANGELOG | match / errata |
| Total bullets | N | M | sum | match / errata |
Phase 2: Analyze (Agent: bkit-impact-analyst)
Input: Phase 1 CC Change Report PLUS Phase 1.5 verification table
Output: bkit Impact Analysis
Pre-flight check before launching the analyst agent: the analyst will
measure bkit architecture stats (agents/skills/hooks counts) via Bash. If
the analyst proposes a numeric correction to memory (e.g. "agents 34 โ 36"),
the main session MUST re-run the measurement command independently before
accepting it. See "Numeric Correction Protocol" in bkit-impact-analyst.md.
Launch the bkit-impact-analyst agent with:
Analyze bkit impact from these CC changes: {phase1_output}
Map each change to bkit components (agents/skills/hooks/lib/scripts).
Identify ENH opportunities starting from ENH-{last+1}.
Check philosophy compliance (Automation First, No Guessing, Docs=Code).
Assess test impact per ENH.
Analysis scope:
- 29 agents: frontmatter compatibility
- 31 skills: allowed-tools, hooks compatibility
- 12 hook events: new events, changed behavior
- 210 lib exports: API compatibility
- 50 scripts: stdin/stdout protocol
- 1,186 TCs: test coverage gaps
Phase 3: Brainstorm (Plan Plus Methodology)
Input: Phase 2 Impact Analysis
Output: Prioritized ENH Roadmap
Apply Plan Plus brainstorming phases:
3.1 Intent Discovery
Ask and answer:
- ์ด CC ์
๊ทธ๋ ์ด๋์์ bkit์ด ์ป์ ์ ์๋ ์ต๋ ๊ฐ์น๋?
- ๋์น๋ฉด ์ ๋๋ critical change๋?
- ๊ธฐ์กด workaround๋ฅผ ๋์ฒดํ ์ ์๋ native ๊ธฐ๋ฅ์?
3.2 Alternative Exploration
For each HIGH/MEDIUM ENH:
- ๊ตฌํ ๋ฐฉ๋ฒ A vs B vs C ๋น๊ต
- ์ต์ ๊ตฌํ (MVP) vs ์์ ๊ตฌํ trade-off
- ๋ค๋ฅธ ENH์์ ์์กด์ฑ/์๋์ง
3.3 YAGNI Review
Each ENH must pass:
- โ
ํ์ฌ ์ฌ์ฉ์๊ฐ ์ค์ ๋ก ํ์๋ก ํ๋๊ฐ?
- โ
๊ตฌํํ์ง ์์ผ๋ฉด ์ด๋ค ๋ฌธ์ ๊ฐ ๋ฐ์ํ๋๊ฐ?
- โ
๋ค์ CC ๋ฒ์ ์์ ๋ ๋์ ๋ฐฉ๋ฒ์ด ๋์ฌ ๊ฐ๋ฅ์ฑ์?
- โ YAGNI fail โ P3 ๊ฐ๋ฑ ๋๋ ์ ๊ฑฐ
3.4 Priority Assignment
Final priority based on:
- P0: Core PDCA workflow ์ง์ ๊ฐ์ ๋๋ known pain point ํด๊ฒฐ
- P1: ์ค์ํ ์ ๊ธฐ๋ฅ ํ์ฑํ ๋๋ major DX ๊ฐ์
- P2: Nice-to-have, ๋ฌธ์ ์
๋ฐ์ดํธ
- P3: Cosmetic, minor optimization, future consideration
Phase 4: Report Generation
Input: All phase outputs
Output: Final Korean report in docs/
- Generate report from
cc-version-analysis.template.md
- Save to:
docs/04-report/features/cc-v{from}-v{to}-impact-analysis.report.md
- Also create Plan (if ENH count > 0):
docs/01-plan/features/cc-v{from}-v{to}-impact-analysis.plan.md
- Update MEMORY.md:
- CC version history section
- ENH number range
- Consecutive compatible releases count
- Open/closed GitHub issues
- Update memory file:
memory/cc_version_history_v{from}_v{to}.md
Task Management Protocol
All work MUST be tracked via Task Management System:
[CC-Version-Analysis] CC v{from} โ v{to} # Parent task
โโโ [Research] Phase 1: CC ๋ณ๊ฒฝ์ฌํญ ์กฐ์ฌ # cc-version-researcher
โ โโโ GitHub releases ์กฐ์ฌ
โ โโโ GitHub issues ์กฐ์ฌ
โ โโโ ๊ณต์ ๋ฌธ์ ๋ณ๊ฒฝ ์กฐ์ฌ
โ โโโ ์์คํ
ํ๋กฌํํธ ๋ณ๊ฒฝ ๋ถ์
โโโ [Analyze] Phase 2: bkit ์ํฅ ๋ถ์ # bkit-impact-analyst
โ โโโ ์ปดํฌ๋ํธ ๋งคํ
โ โโโ ENH ๊ธฐํ ์๋ณ
โ โโโ ํ์ผ ์ํฅ ๋งคํธ๋ฆญ์ค
โ โโโ ์ฒ ํ ์ค์ ๊ฒ์ฆ
โโโ [Brainstorm] Phase 3: ๋ธ๋ ์ธ์คํ ๋ฐ # Plan Plus
โ โโโ ์๋ ํ์
โ โโโ ๋์ ํ์
โ โโโ YAGNI ๊ฒํ
โโโ [Report] Phase 4: ๋ณด๊ณ ์ ์์ฑ # report-generator
โโโ ํ
ํ๋ฆฟ ๊ธฐ๋ฐ ๋ณด๊ณ ์ ์์ฑ
โโโ MEMORY.md ์
๋ฐ์ดํธ
โโโ ์ต์ข
๊ฒํ
Agent Team Configuration
When invoked with CTO Team (/pdca team):
| Role | Agent | Model | Task |
|---|
| Lead | cto-lead | opus | Overall orchestration |
| Researcher | cc-version-researcher | opus | Phase 1: CC research |
| Analyst | bkit-impact-analyst | opus | Phase 2: bkit analysis |
| Reporter | report-generator | haiku | Phase 4: Report writing |
Parallel execution:
- Phase 1 tasks can run in parallel (GitHub, docs, npm)
- Phase 2 depends on Phase 1 completion
- Phase 3 depends on Phase 2 completion
- Phase 4 depends on Phase 3 completion
Quality Checklist
Before completing, verify:
Raw Verification Checklist (NEW โ Phase 1.5 gate, v2.1.16 errata learning)
Known Errata Log (errata learning archive)
When raw verification catches a discrepancy with agent output, record it here
to prevent repeat-mistakes and to feed future skill improvements.
Cycle v2.1.145 (2026-05-20)
| Field | Agent reported | Raw verified | Root cause |
|---|
| Bullet count | 21 (Features 7 + Fixes 13 + Improved 1) | 20 (Added 7 + Fixed 12 + Improved 1) | Agent over-counted Fixed by 1; first WebFetch on model-processed release page under-reported Added by 1 (raw CHANGELOG was authoritative) |
| Agents directory count | 36 (proposed correction from 34) | 34 (ls -1 agents/ | wc -l) | Analyst proposed unverified numeric correction; main session adopted without re-measurement |
| F7-145 background_tasks/session_crons | "extension surface" (no clear source citation) | Confirmed in raw CHANGELOG Added #7 verbatim | Single-source WebFetch had originally omitted this bullet; second raw fetch recovered it |
Lessons applied to skill (this commit):
- Phase 1.5 Raw Verification Gate now mandatory before Phase 2
- cc-version-researcher: verbatim bullet quotation required
- bkit-impact-analyst: direct-measurement-first; Numeric Correction Protocol
- SKILL.md: 7-item Raw Verification Checklist added
- bkit-impact-analyst.md: stale architecture snapshot removed, replaced with mandatory Bash measurement protocol
Previous Analysis Reference
This skill builds on established analysis patterns:
docs/04-report/features/claude-code-v2172-impact-analysis.report.md
docs/04-report/features/claude-code-v2178-impact-analysis.report.md
memory/cc_version_history_v2134_v2172.md
Always read previous reports first to maintain consistency in:
- ENH numbering (continue from last used number)
- Report structure and depth
- Consecutive compatible release tracking
- GitHub issues monitoring continuity