pactkit 2.15.2__tar.gz → 2.16.1__tar.gz

{pactkit-2.15.2 → pactkit-2.16.1}/CHANGELOG.md

@@ -4,6 +4,16 @@ All notable changes to PactKit will be documented in this file.
 
 Format follows [Keep a Changelog](https://keepachangelog.com/).
 
+## [2.16.1] - 2026-07-23
+
+### Fixed
+- **Bedrock VS Code plugin model compatibility** (STORY-slim-134) — Removed `model:` field from all `/project-*` command frontmatter. Claude Code was resolving `model: sonnet/opus` to Anthropic's latest model ID (e.g., `us.anthropic.claude-sonnet-4-5-20250929-v1:0`), bypassing `ANTHROPIC_DEFAULT_SONNET_MODEL` env var in VS Code plugin environments. Commands now inherit the session default model set by the user's provider env vars.
+
+## [2.16.0] - 2026-07-13
+
+### Added
+- **`/project-debug` command** (STORY-slim-133) — Hypothesis-driven troubleshooting skill. Structured loop: Symptom → Hypothesize (≤3) → Verify (executable commands) → Narrow → Root Cause. Enforces evidence-gated file access (no aimless reading) and convergence guarantees (escalates to `/project-plan` if stuck after 3 iterations). Uses sonnet model with structured protocol.
+
 ## [2.15.2] - 2026-07-02
 
 ### Changed

{pactkit-2.15.2 → pactkit-2.16.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pactkit
-Version: 2.15.2
+Version: 2.16.1
 Summary: Spec-driven agentic DevOps toolkit for AI coding assistants
 Project-URL: Homepage, https://pactkit.dev
 Project-URL: Repository, https://github.com/pactkit/pactkit-public

{pactkit-2.15.2 → pactkit-2.16.1}/docs/architecture/governance/archive/lessons_archive_202603.md

@@ -66,3 +66,5 @@
 | 2026-03-26 | DeployerBase static methods use lazy imports to avoid circular dependencies between deploy_base.py and deployer.py | deploy_base.py:DeployerBase.render_prompt |
 | 2026-03-26 | Thin adapter conversion: pactkit-codex reduced from 15,356 lines (full fork) to 668 lines (thin adapter) by importing core via DeployerBase. Key: deploy_codex_playbooks needs explicit path replacement for ALL format prefixes (~/.claude/, ~/.config/opencode/ → ~/.codex/) | pactkit_codex/deployer.py:deploy_codex_playbooks |
 | 2026-03-27 | When migrating commands to skills (subdir/SKILL.md format), all path assertions in pre-existing tests must be updated in the same commit — 17 tests broke because they asserted flat commands/*.md paths | deployer.py:_deploy_commands |
+| 2026-03-30 | Frontend API path convention bugs are structurally preventable: ApiCallParser+api_convention_summary in trace phase surfaces prefix/wrapper conventions before implementation, eliminating a class of copy-paste path errors | visualize.py:ApiCallParser |
+| 2026-03-31 | Rendering-only changes to visualize.py can be isolated by extracting a shared render helper (_render_nested_call_graph) that both forward and reverse BFS call — keeps AST parsing untouched | visualize.py:_render_nested_call_graph |

{pactkit-2.15.2 → pactkit-2.16.1}/docs/architecture/governance/lessons.md

@@ -2,8 +2,6 @@
 
 | Date | Lesson | Context |
 |------|--------|---------|
-| 2026-03-30 | Frontend API path convention bugs are structurally preventable: ApiCallParser+api_convention_summary in trace phase surfaces prefix/wrapper conventions before implementation, eliminating a class of copy-paste path errors | visualize.py:ApiCallParser |
-| 2026-03-31 | Rendering-only changes to visualize.py can be isolated by extracting a shared render helper (_render_nested_call_graph) that both forward and reverse BFS call — keeps AST parsing untouched | visualize.py:_render_nested_call_graph |
 | 2026-03-31 | dict.update() on call_edges causes last-wins overwrite when scanning same-name functions across files; use extend-merge pattern instead | visualize.py:_build_call_graph |
 | 2026-03-31 | CLI args must mirror visualize.py standalone argparse; feature implemented in visualize.py but not exposed in cli.py is effectively dead code | cli.py:viz_parser |
 | 2026-03-31 | tree-sitter comment nodes are direct children of function body nodes (block/statement_block), enabling scoped dispatch hint queries without parent traversal. Per-language comment query needed: Go/TS use (comment), Java uses [(line_comment)(block_comment)]. | visualize.py:_extract_calls_from_body |
@@ -53,3 +51,5 @@
 | 2026-06-10 | Engineering guides are additive: new GUIDES_FILES entries auto-deploy via _deploy_guides() without deployer changes. Test impact limited to count assertions in test_story_slim128. | src/pactkit/prompts/guides.py:GUIDES_FILES |
 | 2026-06-29 | Large @inject rules (architecture 11KB, solution 8.4KB, engineering 4.3KB) removed from COMMAND_RULES_MAP — loaded on-demand via Read in playbooks. Saves ~12,500 tokens/PDCA session. | src/pactkit/prompts/rules.py:COMMAND_RULES_MAP |
 | 2026-07-02 | Hardcoding external tool CLI commands in SKILL_VISUALIZE_MD and deployer.py creates upgrade coupling — replaced with codegraph --help runtime discovery in _build_claude_md_managed_content() | src/pactkit/generators/deployer.py:_build_claude_md_managed_content |
+| 2026-07-13 | Adding project-debug required updating VALID_COMMANDS in config.py, VALID_SKILLS, COMMANDS_CONTENT registration in commands.py, and COMMAND_RULES_MAP in rules.py — plus 6 test files with hardcoded count assertions | src/pactkit/config.py:VALID_COMMANDS, src/pactkit/prompts/rules.py:COMMAND_RULES_MAP |
+| 2026-07-23 | Removing `model:` from COMMANDS_CONTENT frontmatter fixes Bedrock VS Code plugin errors — Claude Code resolves the alias to Anthropic's latest model ID, bypassing `ANTHROPIC_DEFAULT_SONNET_MODEL`; without it, commands inherit the session default model set by the user's env vars | src/pactkit/prompts/commands.py:COMMANDS_CONTENT |

{pactkit-2.15.2 → pactkit-2.16.1}/docs/architecture/governance/rules.md

@@ -15,7 +15,7 @@
 
 ## Invariants
 
-1. All 4032+ tests must pass before any commit to `main`.
+1. All 4045+ tests must pass before any commit to `main`.
 2. Specs (`docs/specs/`) are the source of truth — code conforms to specs, not the reverse.
 3. CLI entry point is `pactkit` via `src/pactkit/cli.py:main`.
 4. No runtime dependencies beyond `pyyaml`.

pactkit-2.16.1/docs/product/archive/archive_202607.md

@@ -0,0 +1,28 @@
+
+### [STORY-slim-132] Decouple codegraph commands from hardcoded prompt
+> Spec: docs/specs/STORY-slim-132.md
+
+- [x] Replace skills.py codegraph block with help fallback
+- [x] Replace deployer.py codegraph section
+- [x] Update global CLAUDE.md
+- [x] Remove MCP tool name list
+
+### [HOTFIX-slim-130] Fix skill frontmatter parsing
+> Spec: docs/specs/HOTFIX-slim-130.md
+
+- [x] Move @ references below YAML frontmatter block in all project-* skills
+
+### [STORY-slim-133] Structured Debug Skill (Hypothesis-Driven Troubleshooting)
+> Spec: docs/specs/STORY-slim-133.md
+
+- [x] Create SKILL.md protocol with all phases
+- [x] Add to routing table in pactkit.md
+- [x] Register in available-skills list
+
+### [STORY-slim-134] remove model field from command frontmatter
+> Spec: docs/specs/STORY-slim-134.md
+
+- [x] 删除 commands.py frontmatter model 字段
+- [x] 删除 workflows.py frontmatter model 字段
+- [x] 运行测试套件
+- [x] redeploy plugin artifact

{pactkit-2.15.2 → pactkit-2.16.1}/docs/product/context.md

@@ -1,8 +1,8 @@
 # Project Context (Auto-generated)
-> Last updated: 2026-07-02T11:00:51+08:00 by pactkit context
+> Last updated: 2026-07-23T18:05:40+08:00 by pactkit context
 
 ## Sprint Status
-Backlog: 4 | In Progress: 0 | Done: 0 stories
+Backlog: 5 | In Progress: 0 | Done: 0 stories
 
 ## Current Stories
 None
@@ -16,11 +16,11 @@ None
 * main
 
 ## Key Decisions
-- codegraph sync must be enforced by code (pactkit visualize/sync), not by prompt instructions — prompt-only enforcement is unreliable for deterministic operations
-- Managed-block pattern (start/end markers + regex replace) is the canonical way to update mixed-ownership files — same pattern for CLAUDE.md and CLAUDE.local.md
 - Engineering guides are additive: new GUIDES_FILES entries auto-deploy via _deploy_guides() without deployer changes. Test impact limited to count assertions in test_story_slim128.
 - Large @inject rules (architecture 11KB, solution 8.4KB, engineering 4.3KB) removed from COMMAND_RULES_MAP — loaded on-demand via Read in playbooks. Saves ~12,500 tokens/PDCA session.
 - Hardcoding external tool CLI commands in SKILL_VISUALIZE_MD and deployer.py creates upgrade coupling — replaced with codegraph --help runtime discovery in _build_claude_md_managed_content()
+- Adding project-debug required updating VALID_COMMANDS in config.py, VALID_SKILLS, COMMANDS_CONTENT registration in commands.py, and COMMAND_RULES_MAP in rules.py — plus 6 test files with hardcoded count assertions
+- Removing `model:` from COMMANDS_CONTENT frontmatter fixes Bedrock VS Code plugin errors — Claude Code resolves the alias to Anthropic's latest model ID, bypassing `ANTHROPIC_DEFAULT_SONNET_MODEL`; without it, commands inherit the session default model set by the user's env vars
 
 ## Next Recommended Action
 `/project-plan`

{pactkit-2.15.2 → pactkit-2.16.1}/docs/product/sprint_board.md

@@ -31,7 +31,6 @@
 
 - [ ] Fix _deploy_commands prepend order
 
-
 ### [HOTFIX-slim-132] Add explicit board.py move_story command to project-act
 > Spec: docs/specs/HOTFIX-slim-132.md
 

pactkit-2.16.1/docs/specs/STORY-slim-125.md

@@ -0,0 +1,72 @@
+# STORY-slim-125: Add model frontmatter to all PDCA command/skill prompts
+
+| Field | Value |
+|-------|-------|
+| ID | STORY-slim-125 |
+| Status | Done |
+| Priority | P1 |
+| Release | 2.14.0 |
+
+## Background
+
+Originally (v2.14.0), all PDCA command frontmatter included a `model:` field to ensure consistent model routing. However, this caused breakage in enterprise Bedrock deployments: Claude Code resolves `model: sonnet` to the latest Anthropic model ID (e.g., `us.anthropic.claude-sonnet-4-5-20250929-v1:0`), which may not be available in a user's Bedrock deployment.
+
+Superseded by STORY-slim-134 (v2.17.0): the `model:` field is removed from all command frontmatter. Commands now inherit the user's session-level default model, which is controlled by environment variables (`ANTHROPIC_DEFAULT_SONNET_MODEL`, etc.) that the user configures for their provider.
+
+## Requirements
+
+### R1: No PDCA command has model frontmatter (MUST)
+
+No command/skill prompt in `src/pactkit/prompts/commands.py` or `src/pactkit/prompts/workflows.py` MUST include a `model:` field in its frontmatter block. Commands inherit the session default model.
+
+### R2: Deploy produces no model field in SKILL.md (MUST)
+
+After `pactkit update`, no deployed `~/.claude/skills/project-*/SKILL.md` MUST contain a `model:` line in its YAML frontmatter.
+
+## Acceptance Criteria
+
+### AC1: No command has model field (R1)
+
+- **Given** the source prompts in `commands.py` and `workflows.py`
+- **When** their frontmatter is parsed
+- **Then** no command prompt contains a `model:` field
+
+### AC2: Deploy produces no model in SKILL.md (R2)
+
+- **Given** `pactkit update` is run
+- **When** deployed SKILL.md files are inspected
+- **Then** no `project-*` skill contains `model:` in frontmatter
+
+### AC3: project-design has no model field (R1)
+
+- **Given** the `project-design` prompt in `workflows.py`
+- **When** its frontmatter is read
+- **Then** no `model:` field is present
+
+## Target Call Chain
+
+```
+commands.py / workflows.py (prompt string with frontmatter)
+  → deployer.py _deploy_skills()
+    → writes ~/.claude/skills/project-*/SKILL.md
+```
+
+## Implementation Steps
+
+| Step | File | Action | Dependencies | Risk |
+|------|------|--------|-------------|------|
+| 1 | `src/pactkit/prompts/commands.py` | Add `model: sonnet` to check, done, clarify, init, release, pr frontmatter | None | Low |
+| 2 | `src/pactkit/prompts/workflows.py` | Add `model: sonnet` to sprint, hotfix; `model: opus` to design | None | Low |
+| 3 | `tests/unit/` | Test that all command prompts have model field | Steps 1-2 | Low |
+
+## Security Scope
+
+| Check | Applicable | Reason |
+|-------|------------|--------|
+| SEC-1 through SEC-8 | N/A | Prompt-only changes, no user input, no external systems |
+
+## Out of Scope
+
+- Changing embedded skill models (pactkit-trace, pactkit-draw, etc.) — already assigned
+- Validating model availability at runtime (Claude Code handles gracefully)
+- Adding model field to agent definitions (separate mechanism via `agent_models` config)

pactkit-2.16.1/docs/specs/STORY-slim-133.md

@@ -0,0 +1,152 @@
+# STORY-slim-133: Structured Debug Skill with Hypothesis-Driven Troubleshooting
+
+| Field | Value |
+|-------|-------|
+| ID | STORY-slim-133 |
+| Status | Done |
+| Priority | P1 |
+| Release | 2.16.0 |
+
+## Background
+
+当前 troubleshooting 流程存在两个核心问题：
+1. **假设发散不收敛** — AI 列出大量可能原因但不逐个验证排除，导致排查停滞
+2. **信息过载** — 漫无目的地 grep/read 大量文件，消耗 token 但不推进定位
+
+现有工具覆盖不了这个场景：
+- `pactkit-trace` 是静态分析工具，前提是"已知要看哪里"
+- `pactkit-doctor` 仅做项目健康检查
+- `system-medic` 仅处理环境/配置问题
+- `project-hotfix` 前提是"已定位到问题"
+
+需要一个**从症状到根因**的结构化排查 skill，强制走「假设→验证→缩小」循环，用 opus 模型的强推理能力替代漫无目的的搜索。
+
+## Requirements
+
+### R1: Hypothesis-Driven Loop (MUST)
+
+Skill MUST enforce a structured loop: Observe → Hypothesize (≤3) → Verify (executable command per hypothesis) → Narrow. Each hypothesis MUST have a concrete verification step (a runnable command or code inspection with expected vs actual outcome). Skill MUST NOT proceed to the next hypothesis without executing verification on the current one first.
+
+### R2: Evidence-Gated File Access (MUST)
+
+Every file read MUST be justified by a stated hypothesis or evidence trail. Skill MUST NOT read files "to get familiar with the codebase" or "to understand the context" without linking to a specific hypothesis being tested — violation of this produces the exact "查不到点子上" problem this skill solves.
+
+### R3: Structured Input Protocol (MUST)
+
+Skill MUST provide a structured input template that minimizes user typing while capturing critical debugging info:
+- **Symptom**: What's happening (error message, unexpected behavior)
+- **Expected**: What should happen instead
+- **Reproduction**: How to trigger it (command, URL, input)
+- **Context**: What changed recently (optional)
+
+When user provides raw text instead of structured input, skill MUST extract and organize into this structure before proceeding.
+
+### R4: Convergence Guarantee (MUST)
+
+Each loop iteration MUST reduce the hypothesis space. If after 3 iterations no hypothesis is confirmed, skill MUST escalate:
+- **Round 1 escalation**: ask user for more information (additional symptoms, reproduction steps)
+- **Round 2 escalation (still stuck)**: nudge to `/project-plan` — "This looks like a cross-module issue that needs full architectural trace."
+
+MUST NOT loop indefinitely on the same hypothesis set. The skill's job is fast triage, not deep surgery.
+
+### R5: Model Selection (MUST)
+
+Skill MUST specify `model: sonnet` in SKILL.md frontmatter. Sonnet + structured protocol handles 80% of common bugs. For complex cross-module issues that don't converge, the escalation path is `/project-plan` (full PDCA with trace), not model upgrade.
+
+### R6: Concluding Report (MUST)
+
+Upon locating root cause, skill MUST output a structured conclusion:
+- **Root Cause**: One-sentence statement
+- **Evidence**: The verification step that confirmed it
+- **Fix Path**: Concrete next action (file:line to change, or command to run)
+- **PDCA Nudge**: Recommend `/project-hotfix` or `/project-act` as appropriate
+
+### R7: Codegraph Integration (SHOULD)
+
+If `.codegraph/` exists, skill SHOULD prefer codegraph for caller/callee analysis over grep during hypothesis verification. Fallback to grep when codegraph is unavailable.
+
+## Acceptance Criteria
+
+### AC1: Hypothesis Loop Enforced (R1, R4)
+
+- **Given** user invokes `/project-debug "tests fail with KeyError on user_id"`
+- **When** the skill starts processing
+- **Then** it outputs: (1) structured symptom summary, (2) ≤3 ranked hypotheses each with a verification command, (3) executes verification commands in order, (4) narrows based on results, (5) converges to root cause within ≤5 iterations
+
+### AC2: No Aimless File Reading (R2)
+
+- **Given** the skill is in a verification phase
+- **When** it needs to read a file
+- **Then** the read is preceded by a one-line justification linking to a hypothesis (e.g., "H2 verification: checking if `models.py:45` validates user_id before access")
+
+### AC3: Structured Input Extraction (R3)
+
+- **Given** user provides unstructured text like "部署之后接口报500了"
+- **When** the skill receives this input
+- **Then** it extracts and formats into the 4-field template, asks for missing critical fields (Reproduction), and does NOT start hypothesizing until Symptom is clear
+
+### AC4: Convergence Escalation (R4)
+
+- **Given** 3 iterations have passed without confirming a hypothesis
+- **When** the skill enters the 4th iteration
+- **Then** it asks user for more info; if still stuck after 1 more round, it nudges `/project-plan` with a summary of what was tried and ruled out
+
+### AC5: Sonnet Model with PDCA Escalation (R5)
+
+- **Given** the SKILL.md file exists at `~/.claude/skills/project-debug/SKILL.md`
+- **When** inspecting the frontmatter
+- **Then** the `model` field is set to `sonnet`
+
+### AC6: Conclusion Report Format (R6)
+
+- **Given** root cause has been identified through the verification loop
+- **When** the skill outputs its conclusion
+- **Then** it includes all 4 fields: Root Cause (one sentence), Evidence (which verification confirmed it), Fix Path (file:line or command), and PDCA Nudge (appropriate command recommendation)
+
+### AC7: Codegraph Preferred When Available (R7)
+
+- **Given** the project has `.codegraph/` directory
+- **When** the skill needs to trace callers/callees during verification
+- **Then** it uses `codegraph` CLI instead of raw grep
+
+## Target Call Chain
+
+```
+User invokes /project-debug "$symptom"
+→ SKILL.md protocol loaded (model: opus)
+→ Phase 1: Symptom Structuring (extract/confirm 4-field template)
+→ Phase 2: Hypothesis Generation (≤3, ranked by probability)
+→ Phase 3: Verification Loop (execute → observe → narrow)
+  └─ Loop until: root cause confirmed OR escalation triggered
+→ Phase 4: Conclusion Report (root cause + fix path + PDCA nudge)
+```
+
+No code implementation needed — this is a pure SKILL.md protocol (prompt-as-code for diagnostic reasoning).
+
+## Implementation Steps
+
+| Step | File | Action | Dependencies | Risk |
+|------|------|--------|-------------|------|
+| 1 | `~/.claude/skills/project-debug/SKILL.md` | Create skill protocol with all phases | None | Low |
+| 2 | `~/.claude/rules/pactkit.md` | Add command to routing table | Step 1 | Low |
+| 3 | Board | Register skill in available-skills | Step 1 | Low |
+
+## Security Scope
+
+| Check | Applicable | Reason |
+|-------|------------|--------|
+| SEC-1 (Injection) | N/A | No user input passed to shell unsanitized — verification commands are AI-composed |
+| SEC-2 (Auth) | N/A | No auth logic involved |
+| SEC-3 (Data Exposure) | N/A | Reads local files only |
+| SEC-4 (XSS) | N/A | No web output |
+| SEC-5 (Config) | N/A | No config files modified |
+| SEC-6 (Crypto) | N/A | No crypto involved |
+| SEC-7 (Logging) | N/A | No logging changes |
+| SEC-8 (SSRF) | N/A | No network requests |
+
+## Out of Scope
+
+- 自动修复代码 — skill 仅定位问题，修复走 `/project-hotfix` 或 `/project-act`
+- 环境/配置问题 — 已有 `system-medic` + `pactkit-doctor` 覆盖
+- 性能分析 — 属于另一个专项 skill 的范畴
+- CI/CD pipeline 问题 — 超出本地 debug 范围

pactkit-2.16.1/docs/specs/STORY-slim-134.md

@@ -0,0 +1,102 @@
+# STORY-slim-134: remove model field from command frontmatter
+
+| Field | Value |
+|-------|-------|
+| ID | STORY-slim-134 |
+| Status | Done |
+| Priority | P1 |
+| Release | 2.17.0 |
+
+## Background
+
+Claude Code 在读取 command frontmatter 中的 `model: sonnet` / `model: opus` 字段时，会将 alias 解析成 Anthropic 官方最新的 model ID（如 `us.anthropic.claude-sonnet-4-5-20250929-v1:0`）。但企业用户通过 Bedrock 代理（如 genai-nexus）接入时，其部署往往只启用了较旧版本的模型，导致 VS Code 插件环境下所有 `/project-*` 命令都因模型不可用而报错。
+
+移除 frontmatter 中的 `model:` 字段后，command 会继承用户当前 session 的默认模型（由用户的 `ANTHROPIC_DEFAULT_SONNET_MODEL` 等环境变量或 Claude Code 设置控制），不再主动切换，从而避免 Bedrock 兼容性问题。
+
+## Requirements
+
+### R1: 移除 prompts/commands.py 中的 model 字段 (MUST)
+
+删除 `src/pactkit/prompts/commands.py` 中所有 command 字符串 frontmatter 里的 `model:` 行。涉及 9 处（`model: opus` × 2，`model: sonnet` × 7）。
+
+### R2: 移除 prompts/workflows.py 中的 model 字段 (MUST)
+
+删除 `src/pactkit/prompts/workflows.py` 中所有 command 字符串 frontmatter 里的 `model:` 行。涉及 4 处（`model: opus` × 1，`model: sonnet` × 3）。
+
+### R3: 同步更新 pactkit-plugin/commands/ (MUST)
+
+执行 `pactkit deploy --format plugin` 重新生成 `pactkit-plugin/commands/*.md`，确保 generated artifact 与 source 一致。
+
+### R4: 不影响 sprint 内联 model 引用 (MUST NOT)
+
+`workflows.py` 中 sprint 命令正文里对 `model: sonnet` 的**文本说明**（如 `A1 model: opus`）不得修改——这些是给 LLM 读的 prompt 指令，不是 frontmatter 字段。
+
+## Acceptance Criteria
+
+### AC1: commands.py 无 model 字段 (R1)
+
+- **Given** `src/pactkit/prompts/commands.py` 中各 command 字符串的 YAML frontmatter
+- **When** 检查每个 `---` 块内的字段列表
+- **Then** 不存在任何 `model:` 行；`description:` 和 `allowed-tools:` 字段保持不变
+
+### AC2: workflows.py 无 model 字段 (R2)
+
+- **Given** `src/pactkit/prompts/workflows.py` 中各 command 字符串的 YAML frontmatter
+- **When** 检查每个 `---` 块内的字段列表
+- **Then** 不存在任何 `model:` 行；正文中对模型的文字引用保持不变
+
+### AC3: plugin artifact 与 source 一致 (R3)
+
+- **Given** R1/R2 修改完成后执行 `pactkit deploy --format plugin`
+- **When** 检查 `pactkit-plugin/commands/*.md` 各文件的 frontmatter
+- **Then** 所有文件 frontmatter 中均无 `model:` 字段
+
+### AC5: sprint 正文模型引用不变 (R4)
+
+- **Given** R1/R2 修改完成
+- **When** 检查 `workflows.py` 中 sprint 命令正文（非 frontmatter）的内容
+- **Then** `A1 (system-architect, model: opus)`、`A2 (senior-developer, model: sonnet)` 等文字说明保持不变
+
+### AC4: 测试套件全通过 (R1-R4)
+
+- **Given** 修改完成
+- **When** 运行 `.venv/bin/pytest tests/ -v`
+- **Then** 所有测试通过，无新失败
+
+## Target Call Chain
+
+```
+src/pactkit/prompts/commands.py (COMMANDS_CONTENT 字符串常量，含 model: 字段)
+src/pactkit/prompts/workflows.py (SPRINT_PROMPT / HOTFIX_PROMPT 等，含 model: 字段)
+  → deployer._deploy_commands() — 原样写出字符串到文件
+  → pactkit-plugin/commands/*.md (generated artifact)
+  → Claude Code frontmatter 解析 → model alias → Bedrock model ID
+```
+
+## Implementation Steps
+
+| Step | File | Action | Dependencies | Risk |
+|------|------|--------|-------------|------|
+| 1 | `src/pactkit/prompts/commands.py` | 删除所有 frontmatter 中的 `model:` 行（9 处） | None | Low |
+| 2 | `src/pactkit/prompts/workflows.py` | 删除所有 frontmatter 中的 `model:` 行（4 处） | Step 1 | Low |
+| 3 | 运行测试 | `.venv/bin/pytest tests/ -v` 确保无回归 | Step 1-2 | Low |
+| 4 | `pactkit-plugin/commands/` | `pactkit deploy --format plugin` 重新生成 artifact | Step 1-2 | Low |
+
+## Security Scope
+
+| Check | Applicable | Reason |
+|-------|------------|--------|
+| SEC-1 | N/A | 仅删除字符串字段，无新代码逻辑引入 |
+| SEC-2 | N/A | 无输入处理变更 |
+| SEC-3 | N/A | 无数据库相关 |
+| SEC-4 | N/A | 无前端文件 |
+| SEC-5 | N/A | 无认证/会话逻辑变更 |
+| SEC-6 | N/A | 无 API/路由变更 |
+| SEC-7 | N/A | 无错误处理变更 |
+| SEC-8 | N/A | 无依赖变更 |
+
+## Out of Scope
+
+- 不修改 `deployer.py` 中的 frontmatter 处理逻辑
+- 不为用户提供 model fallback 配置机制（该需求已明确放弃）
+- 不修改 sprint/hotfix 正文中对模型的文字说明（R4）

{pactkit-2.15.2 → pactkit-2.16.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "pactkit"
-version = "2.15.2"
+version = "2.16.1"
 description = "Spec-driven agentic DevOps toolkit for AI coding assistants"
 readme = "README.md"
 license = "MIT"

{pactkit-2.15.2 → pactkit-2.16.1}/src/pactkit/__init__.py

@@ -1,3 +1,3 @@
 """PactKit - Spec-driven agentic DevOps toolkit."""
 
-__version__ = "2.15.2"
+__version__ = "2.16.1"

{pactkit-2.15.2 → pactkit-2.16.1}/src/pactkit/config.py

@@ -40,6 +40,7 @@ VALID_COMMANDS = frozenset(
         "project-clarify",
         "project-release",
         "project-pr",
+        "project-debug",
     }
 )
 
@@ -71,6 +72,7 @@ VALID_SKILLS = frozenset(
         "project-clarify",
         "project-release",
         "project-pr",
+        "project-debug",
     }
 )
 

{pactkit-2.15.2 → pactkit-2.16.1}/src/pactkit/prompts/commands.py

@@ -1,4 +1,5 @@
 from pactkit.prompts.workflows import (
+    DEBUG_PROMPT,
     DESIGN_PROMPT,
     HOTFIX_PROMPT,
     SPRINT_PROMPT,
@@ -8,7 +9,6 @@ COMMANDS_CONTENT = {
     "project-plan.md": """---
 description: "Analyze requirements, create Spec and Story"
 allowed-tools: [Read, Write, Edit, Bash, Glob, Grep]
-model: opus
 ---
 
 # Command: Plan (v1.3.0 Integrated Trace)
@@ -166,7 +166,6 @@ model: opus
     "project-act.md": """---
 description: "Implement code per Spec, strict TDD"
 allowed-tools: [Read, Write, Edit, Bash, Glob, Grep]
-model: sonnet
 ---
 
 # Command: Act (v1.3.0 Stack-Aware)
@@ -287,7 +286,6 @@ model: sonnet
     "project-check.md": """---
 description: "QA verification: security scan, code quality scan, Spec alignment"
 allowed-tools: [Read, Bash, Grep, Glob]
-model: sonnet
 ---
 
 # Command: Check (v1.3.0 Deep QA)
@@ -501,7 +499,6 @@ For each finding, assign a severity (P0-P3). Flag issues that may cause silent f
     "project-done.md": """---
 description: "Code cleanup, Board update, Git commit"
 allowed-tools: [Read, Write, Edit, Bash, Glob]
-model: sonnet
 ---
 
 # Command: Done (v1.3.0 Smart Gatekeeper)
@@ -672,7 +669,6 @@ Run `pactkit coverage-gate <changed-files>` to verify coverage on changed source
     "project-clarify.md": """---
 description: "Standalone requirement clarification before planning"
 allowed-tools: [Read, Bash, Glob, Grep]
-model: sonnet
 ---
 
 # Command: Clarify (v1.1.0)
@@ -713,7 +709,6 @@ model: sonnet
     "project-init.md": """---
 description: "Initialize project scaffolding and governance structure"
 allowed-tools: [Read, Write, Edit, Bash, Glob]
-model: sonnet
 ---
 
 # Command: Init (v1.3.0 Rich)
@@ -786,7 +781,6 @@ model: sonnet
     "project-release.md": """---
 description: "Version release: snapshot, archive, Git tag, and GitHub Release"
 allowed-tools: [Read, Write, Edit, Bash, Glob]
-model: sonnet
 ---
 
 # Command: Release (v1.4.0)
@@ -809,7 +803,6 @@ model: sonnet
     "project-pr.md": """---
 description: "Push branch and create pull request via gh CLI"
 allowed-tools: [Read, Write, Edit, Bash, Glob]
-model: sonnet
 ---
 
 # Command: PR (v1.4.0)
@@ -868,3 +861,4 @@ model: sonnet
 COMMANDS_CONTENT["project-sprint.md"] = SPRINT_PROMPT
 COMMANDS_CONTENT["project-hotfix.md"] = HOTFIX_PROMPT
 COMMANDS_CONTENT["project-design.md"] = DESIGN_PROMPT
+COMMANDS_CONTENT["project-debug.md"] = DEBUG_PROMPT

{pactkit-2.15.2 → pactkit-2.16.1}/src/pactkit/prompts/rules.py

@@ -857,6 +857,7 @@ COMMAND_RULES_MAP = {
         "core", "sectional", "hierarchy", "atlas", "routing", "workflow",
         "mcp", "shared", "credential",
     ],
+    "project-debug": ["core", "shared", "credential"],
 }
 
 # Managed file prefixes for rules/ directory cleanup (deployer will clean these, leave user files intact)