@laitszkin/apollo-toolkit 3.12.0 → 3.12.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/AGENTS.md +6 -6
- package/CHANGELOG.md +7 -4
- package/README.md +9 -10
- package/analyse-app-logs/scripts/__pycache__/filter_logs_by_time.cpython-312.pyc +0 -0
- package/analyse-app-logs/scripts/__pycache__/log_cli_utils.cpython-312.pyc +0 -0
- package/analyse-app-logs/scripts/__pycache__/search_logs.cpython-312.pyc +0 -0
- package/commit-and-push/SKILL.md +1 -3
- package/docs-to-voice/scripts/__pycache__/docs_to_voice.cpython-312.pyc +0 -0
- package/enhance-existing-features/SKILL.md +21 -37
- package/generate-spec/SKILL.md +7 -10
- package/generate-spec/scripts/__pycache__/create-specscpython-312.pyc +0 -0
- package/init-project-html/SKILL.md +15 -17
- package/iterative-code-performance/SKILL.md +1 -1
- package/iterative-code-quality/SKILL.md +1 -1
- package/katex/scripts/__pycache__/render_katex.cpython-312.pyc +0 -0
- package/maintain-project-constraints/SKILL.md +18 -22
- package/merge-changes-from-local-branches/SKILL.md +23 -34
- package/open-github-issue/scripts/__pycache__/open_github_issue.cpython-312.pyc +0 -0
- package/open-source-pr-workflow/SKILL.md +4 -7
- package/optimise-skill/SKILL.md +8 -8
- package/optimise-skill/references/definition.md +1 -0
- package/optimise-skill/references/example_skill.md +8 -8
- package/package.json +1 -1
- package/read-github-issue/scripts/__pycache__/find_issues.cpython-312.pyc +0 -0
- package/read-github-issue/scripts/__pycache__/read_issue.cpython-312.pyc +0 -0
- package/resolve-review-comments/scripts/__pycache__/review_threads.cpython-312.pyc +0 -0
- package/review-spec-related-changes/SKILL.md +30 -38
- package/ship-github-issue-fix/SKILL.md +2 -2
- package/solve-issues-found-during-review/SKILL.md +8 -43
- package/spec-to-project-html/SKILL.md +2 -2
- package/submission-readiness-check/SKILL.md +3 -19
- package/systematic-debug/SKILL.md +2 -2
- package/text-to-short-video/scripts/__pycache__/enforce_video_aspect_ratio.cpython-312.pyc +0 -0
- package/version-release/SKILL.md +3 -3
- package/discover-edge-cases/CHANGELOG.md +0 -19
- package/discover-edge-cases/LICENSE +0 -21
- package/discover-edge-cases/README.md +0 -87
- package/discover-edge-cases/SKILL.md +0 -32
- package/discover-edge-cases/agents/openai.yaml +0 -4
- package/discover-edge-cases/references/architecture-edge-cases.md +0 -41
- package/discover-edge-cases/references/code-edge-cases.md +0 -46
- package/discover-security-issues/CHANGELOG.md +0 -32
- package/discover-security-issues/LICENSE +0 -21
- package/discover-security-issues/README.md +0 -35
- package/discover-security-issues/SKILL.md +0 -54
- package/discover-security-issues/agents/openai.yaml +0 -4
- package/discover-security-issues/references/agent-attack-catalog.md +0 -117
- package/discover-security-issues/references/common-software-attack-catalog.md +0 -168
- package/discover-security-issues/references/red-team-extreme-scenarios.md +0 -81
- package/discover-security-issues/references/risk-checklist.md +0 -78
- package/discover-security-issues/references/security-test-patterns-agent.md +0 -101
- package/discover-security-issues/references/security-test-patterns-finance.md +0 -88
- package/discover-security-issues/references/test-snippets.md +0 -73
- package/recover-missing-plan/SKILL.md +0 -85
- package/recover-missing-plan/agents/openai.yaml +0 -4
- package/review-change-set/LICENSE +0 -21
- package/review-change-set/README.md +0 -55
- package/review-change-set/SKILL.md +0 -46
- package/review-change-set/agents/openai.yaml +0 -4
- package/review-codebases/LICENSE +0 -21
- package/review-codebases/README.md +0 -69
- package/review-codebases/SKILL.md +0 -46
- package/review-codebases/agents/openai.yaml +0 -4
- package/scheduled-runtime-health-check/LICENSE +0 -21
- package/scheduled-runtime-health-check/README.md +0 -107
- package/scheduled-runtime-health-check/SKILL.md +0 -135
- package/scheduled-runtime-health-check/agents/openai.yaml +0 -4
- package/scheduled-runtime-health-check/references/output-format.md +0 -20
package/AGENTS.md
CHANGED
|
@@ -31,10 +31,10 @@ This repository enables users to install and run a curated set of reusable agent
|
|
|
31
31
|
- Users can turn lecture slides, past papers, and answer books into mock exams, worked solutions, study notes, or graded PDFs with KaTeX-rendered math.
|
|
32
32
|
- Users can extend existing features in a brownfield codebase with required tests and approvals.
|
|
33
33
|
- Users can propose product features from an existing codebase and publish accepted proposals.
|
|
34
|
-
|
|
34
|
+
|
|
35
35
|
- Users can read, filter, and inspect remote GitHub issues before planning follow-up work.
|
|
36
36
|
- Users can resolve a GitHub issue end-to-end and push the fix directly to a requested branch without opening a PR.
|
|
37
|
-
|
|
37
|
+
|
|
38
38
|
- Users can run a shared submission-readiness pass that synchronizes changelog, project docs, `AGENTS.md`, and completed plan archives before commit, push, PR creation, or release.
|
|
39
39
|
- Users can learn new or improved skills from recent Codex conversation history.
|
|
40
40
|
- Users can audit and maintain the skill catalog itself, including dependency classification and shared-skill extraction decisions.
|
|
@@ -51,12 +51,12 @@ This repository enables users to install and run a curated set of reusable agent
|
|
|
51
51
|
- Users can generate storyboard image sets from chapters, novels, articles, or scripts.
|
|
52
52
|
- Users can configure OpenClaw from official documentation, including `~/.openclaw/openclaw.json`, skills loading, SecretRefs, CLI edits, and validation or repair workflows.
|
|
53
53
|
- Users can record multi-account spending and balance changes in monthly Excel ledgers with summary analytics and charts.
|
|
54
|
-
|
|
55
|
-
|
|
54
|
+
|
|
55
|
+
|
|
56
56
|
- Users can review recent or user-specified spec-backed changes against the governing planning documents, treating unmet business goals as the most severe findings before secondary edge-case, security, and code-review checks.
|
|
57
57
|
- Users can process GitHub pull request review comments and resolve addressed threads.
|
|
58
|
-
|
|
59
|
-
|
|
58
|
+
|
|
59
|
+
|
|
60
60
|
- Users can investigate gated or shadow LLM APIs by capturing real client request shapes, replaying verified traffic patterns, and attributing the likely underlying model through black-box fingerprinting.
|
|
61
61
|
- Users can build and maintain Solana programs and Rust clients using official Solana development workflows.
|
|
62
62
|
- Users can fix issues discovered during a review pass by processing findings from highest to lowest severity, with per-fix validation and full-scope re-validation.
|
package/CHANGELOG.md
CHANGED
|
@@ -2,13 +2,16 @@
|
|
|
2
2
|
|
|
3
3
|
All notable changes to this repository are documented in this file.
|
|
4
4
|
|
|
5
|
-
## [
|
|
6
|
-
|
|
7
|
-
### Added
|
|
5
|
+
## [v3.12.1] - 2026-05-14
|
|
8
6
|
|
|
9
7
|
### Changed
|
|
10
8
|
|
|
11
|
-
|
|
9
|
+
- Rewrite remaining shipped skills (`enhance-existing-features`, `generate-spec`, `init-project-html`, `merge-changes-from-local-branches`, `open-source-pr-workflow`, `optimise-skill`, `review-spec-related-changes`, `solve-issues-found-during-review`, `submission-readiness-check`, `commit-and-push`, `version-release`, `iterative-code-performance`, `iterative-code-quality`, `maintain-project-constraints`, `ship-github-issue-fix`, `spec-to-project-html`, `systematic-debug`) into the compact Chinese-first structure with goal, acceptance criteria, workflow, examples, and references sections.
|
|
10
|
+
- Update `AGENTS.md` and `README.md` to remove references to deleted skills and simplify dependency descriptions.
|
|
11
|
+
|
|
12
|
+
### Removed
|
|
13
|
+
|
|
14
|
+
- Remove `discover-edge-cases`, `discover-security-issues`, `recover-missing-plan`, `review-change-set`, `review-codebases`, and `scheduled-runtime-health-check` skills that are no longer part of the curated catalog.
|
|
12
15
|
|
|
13
16
|
## [v3.12.0] - 2026-05-13
|
|
14
17
|
|
package/README.md
CHANGED
|
@@ -12,7 +12,7 @@ A curated skill catalog for Codex, OpenClaw, Trae, Agents, and Claude Code with
|
|
|
12
12
|
- codex-memory-manager
|
|
13
13
|
- deep-research-topics
|
|
14
14
|
- develop-new-features
|
|
15
|
-
|
|
15
|
+
|
|
16
16
|
- docs-to-voice
|
|
17
17
|
- exam-pdf-workflow
|
|
18
18
|
- document-vision-reader
|
|
@@ -21,7 +21,7 @@ A curated skill catalog for Codex, OpenClaw, Trae, Agents, and Claude Code with
|
|
|
21
21
|
- financial-research
|
|
22
22
|
- read-github-issue
|
|
23
23
|
- generate-spec
|
|
24
|
-
|
|
24
|
+
|
|
25
25
|
- implement-specs
|
|
26
26
|
- implement-specs-with-subagents
|
|
27
27
|
- implement-specs-with-worktree
|
|
@@ -41,13 +41,13 @@ A curated skill catalog for Codex, OpenClaw, Trae, Agents, and Claude Code with
|
|
|
41
41
|
- openai-text-to-image-storyboard
|
|
42
42
|
- openclaw-configuration
|
|
43
43
|
- optimise-skill
|
|
44
|
-
|
|
44
|
+
|
|
45
45
|
- record-spending
|
|
46
46
|
- resolve-review-comments
|
|
47
|
-
|
|
48
|
-
|
|
47
|
+
|
|
48
|
+
|
|
49
49
|
- review-spec-related-changes
|
|
50
|
-
|
|
50
|
+
|
|
51
51
|
- shadow-api-model-research
|
|
52
52
|
- solana-development
|
|
53
53
|
- spec-to-project-html
|
|
@@ -208,12 +208,11 @@ Compatibility note:
|
|
|
208
208
|
|
|
209
209
|
- `generate-spec` is a local skill used by `develop-new-features` and `enhance-existing-features`, and it can produce either a single spec set under `docs/plans/{YYYY-MM-DD}/{change_name}/` or a coordinated parallel batch under `docs/plans/{YYYY-MM-DD}/{batch_name}/{change_name}/` with shared `coordination.md`.
|
|
210
210
|
- `implement-specs-with-subagents` coordinates one `implement-specs-with-worktree` subagent per spec directory for approved multi-spec batches.
|
|
211
|
-
|
|
211
|
+
|
|
212
212
|
- `read-github-issue` uses GitHub CLI (`gh`) directly for remote issue discovery and inspection, so it does not add any extra skill dependency.
|
|
213
|
-
- `review-spec-related-changes` is a local skill that
|
|
213
|
+
- `review-spec-related-changes` is a local skill that reviews spec compliance of changes against governing planning documents, assessing business goals before secondary code-practice concerns.
|
|
214
214
|
- `update-project-html` is a local skill that depends on `init-project-html` for semantic rules and on the `apltk architecture` CLI to refresh the base atlas after code changes; for spec overlay diagrams use `spec-to-project-html` instead.
|
|
215
|
-
|
|
216
|
-
- `review-change-set` no longer chains `discover-security-issues`; for multi-file diffs it prefers dispatching one read-only subagent per scope cluster.
|
|
215
|
+
|
|
217
216
|
|
|
218
217
|
## Release publishing
|
|
219
218
|
|
|
Binary file
|
|
Binary file
|
|
Binary file
|
package/commit-and-push/SKILL.md
CHANGED
|
@@ -19,9 +19,7 @@ description: 提供提交指引以及作為提交前的必要品控閘門。當
|
|
|
19
19
|
|
|
20
20
|
### 2. 品控閘門(選用)
|
|
21
21
|
|
|
22
|
-
|
|
23
|
-
如果外部環境允許使用subagents,建議通過調度subagents完成三個不同維度的代碼審查工作。
|
|
24
|
-
如果在審查過程中發現問題,修復發現的問題並暫存。
|
|
22
|
+
如果變更範圍涉及代碼變更,確認變更已通過必要的審查與驗證。如果在審查過程中發現問題,修復發現的問題並暫存。
|
|
25
23
|
|
|
26
24
|
### 3. 同步項目文檔
|
|
27
25
|
|
|
Binary file
|
|
@@ -1,56 +1,40 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: enhance-existing-features
|
|
3
3
|
description: >-
|
|
4
|
-
|
|
5
|
-
還是先走 `generate-spec` / `recover-missing-plan`;所有非平凡變更都要經過 `test-case-strategy`。
|
|
4
|
+
在既有系統上增強或調整產品行為。先讀懂受影響模組,再決定直接實作或走 generate-spec;所有非平凡變更必須經過 test-case-strategy。
|
|
6
5
|
---
|
|
7
6
|
|
|
8
|
-
# 增強既有功能
|
|
9
|
-
|
|
10
|
-
## Dependencies
|
|
11
|
-
|
|
12
|
-
- Required: `test-case-strategy`
|
|
13
|
-
- Conditional: `generate-spec`、`recover-missing-plan`、`commit-and-push`
|
|
14
|
-
- Optional: 無
|
|
15
|
-
- Fallback: `test-case-strategy` 不可用時必須停止;需要spec但 `generate-spec` 或 `recover-missing-plan` 不可用時也必須停止;若使用者要求提交或推送但 `commit-and-push` 不可用,必須說明後停止
|
|
16
|
-
|
|
17
|
-
## Standards
|
|
18
|
-
|
|
19
|
-
- Evidence: 先閱讀實際程式碼、整合點與外部文件,再決定流程與修改方案
|
|
20
|
-
- Execution: 探索範圍 -> 判斷是否需要 spec -> 查官方資料 -> 最小實作 -> 測試 -> 回填或總結
|
|
21
|
-
- Quality: 不因怕麻煩而跳過 spec,也不為小改動硬開 spec;只做與需求直接相關的修改
|
|
22
|
-
- Output: 交付符合需求的行為變更、可追溯的測試結果,以及在需要時保持誠實的計劃文件狀態
|
|
23
|
-
|
|
24
7
|
## 技能目標
|
|
25
8
|
|
|
26
|
-
在不打亂既有系統邊界的前提下,為 brownfield
|
|
9
|
+
在不打亂既有系統邊界的前提下,為 brownfield 專案安全地增強功能或調整行為,以合適的 spec 流程與測試策略將風險控制在可驗證範圍內。
|
|
27
10
|
|
|
28
11
|
## 驗收條件
|
|
29
12
|
|
|
30
|
-
-
|
|
31
|
-
-
|
|
32
|
-
- 每個非平凡變更都經過 `test-case-strategy
|
|
33
|
-
-
|
|
13
|
+
- 變更範圍明確:已讀懂受影響模組、入口點、整合邊界與變更爆炸半徑。
|
|
14
|
+
- 已正確判斷直接實作或必須先走 `generate-spec`,並在不需要 spec 時給出明確理由。
|
|
15
|
+
- 每個非平凡變更都經過 `test-case-strategy`,留下清楚的 oracle、驗證方式與通過證據。
|
|
16
|
+
- 若使用 spec:批准、實作、回填與計劃文件狀態同步全部完成。
|
|
17
|
+
- 若未使用 spec:最終摘要足以說明風險、測試結果與剩餘限制。
|
|
18
|
+
- `test-case-strategy` 不可用時必須停止;需要 spec 但 `generate-spec` 不可用時必須停止。
|
|
34
19
|
|
|
35
20
|
## 工作流程
|
|
36
21
|
|
|
37
|
-
1.
|
|
38
|
-
2. 判斷是否需要spec
|
|
39
|
-
3.
|
|
40
|
-
4.
|
|
41
|
-
5.
|
|
42
|
-
6.
|
|
43
|
-
7. 若使用了 spec,回填 `spec.md`、`tasks.md`、`checklist.md`、`contract.md`、`design.md` 與任何 `coordination.md`;若未使用 spec,輸出簡潔的結果摘要、測試證據與剩餘限制
|
|
22
|
+
1. 完整閱讀受影響模組、資料流、整合點與現有測試,明確界定變更的爆炸半徑。
|
|
23
|
+
2. 判斷是否需要 spec:涉及新的或改變使用者可見行為、跨模組協調、高風險敏感流程、重大歧義時走 `generate-spec`。不需 spec 時明確說出理由。
|
|
24
|
+
3. 若需要 spec,必須等批准後才能改產品程式碼。
|
|
25
|
+
4. 查閱變更依賴的官方文件或 API spec,以最小差異實作需求,不順手做額外重構或範圍擴張。
|
|
26
|
+
5. 使用 `test-case-strategy` 選擇測試組合,跑測試並修正失敗。
|
|
27
|
+
6. 若使用 spec,回填所有計劃文件;若未使用,輸出簡潔的結果摘要、測試證據與剩餘限制。
|
|
44
28
|
|
|
45
29
|
## 使用範例
|
|
46
30
|
|
|
47
|
-
- 「新增一套影響 API、資料庫與 UI
|
|
48
|
-
- 「修正分頁 off-by-one,讓行為回到 README
|
|
49
|
-
|
|
31
|
+
- 「新增一套影響 API、資料庫與 UI 的權限模型」→ 先走 `generate-spec`,批准後再實作,並補上跨層測試。交付物包含已批准的 spec、通過的測試證據、回填完成的計劃文件。
|
|
32
|
+
- 「修正分頁 off-by-one,讓行為回到 README 描述」→ 不開 spec,直接修復並加回歸測試。交付物為修復後的代碼與測試通過證據。
|
|
33
|
+
|
|
50
34
|
|
|
51
35
|
## 參考資料索引
|
|
52
36
|
|
|
53
|
-
- `generate-spec`:需要書面批准時的spec
|
|
54
|
-
|
|
55
|
-
- `test-case-strategy`:非平凡變更的測試選型與 oracle
|
|
56
|
-
- `commit-and-push
|
|
37
|
+
- `generate-spec`:需要書面批准時的 spec 流程。
|
|
38
|
+
|
|
39
|
+
- `test-case-strategy`:非平凡變更的測試選型與 oracle 設計。
|
|
40
|
+
- `commit-and-push`:使用者要求提交或推送時的最終交付流程。
|
package/generate-spec/SKILL.md
CHANGED
|
@@ -1,9 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: generate-spec
|
|
3
|
-
description:
|
|
4
|
-
Create or refresh approval-gated planning docs under `docs/plans/...` with
|
|
5
|
-
`apltk create-specs`, `test-case-strategy`, and optional batch coordination or
|
|
6
|
-
preparation files. Use for drafting or restructuring specs, not implementation.
|
|
3
|
+
description: 當你需要將用戶模糊的複雜需求拆解成有嚴格實作範圍的spec時,使用這個技能
|
|
7
4
|
---
|
|
8
5
|
|
|
9
6
|
## 目標
|
|
@@ -15,8 +12,8 @@ description: >-
|
|
|
15
12
|
|
|
16
13
|
## 工作流程
|
|
17
14
|
1. 理解用戶需求並閱讀repo
|
|
18
|
-
分析用戶需求,並在repo之中搜索、列出可能相關的內容。完成搜索之後,深入閱讀相關代碼,識別變更範圍。
|
|
19
|
-
如果外部環境存在subagents功能,建議通過調度subagents來完成深入閱讀repo的任務。
|
|
15
|
+
分析用戶需求,並在 repo 之中搜索、列出可能相關的內容。完成搜索之後,深入閱讀相關代碼,識別變更範圍。
|
|
16
|
+
如果外部環境存在 subagents 功能,建議通過調度 subagents 來完成深入閱讀 repo 的任務。
|
|
20
17
|
|
|
21
18
|
2. 拆分用戶需求及設計業務架構
|
|
22
19
|
將用戶需求轉化、拆分為明確、存在邊界的工程需求。結合現有代碼,設計業務架構。在設計的過程中,你需要考慮包括但不限於以下設計事項:
|
|
@@ -33,13 +30,13 @@ description: >-
|
|
|
33
30
|
為任務制定基於測試的驗收條件,確保每一個任務在完成之後都能夠被驗證。
|
|
34
31
|
同時,為需求制定驗收條件,確保用戶需求能夠被測試清晰地驗收、檢驗成果。
|
|
35
32
|
|
|
36
|
-
5. 使用 `apltk` cli工具協助完成spec
|
|
37
|
-
使用cli工具,產生spec的模板。將你的完整計劃填入到模板之中,並通過cli
|
|
38
|
-
如果該spec設計超過三個模塊,則需要創建batch spec。
|
|
33
|
+
5. 使用 `apltk` cli工具協助完成spec及spec相關架構圖
|
|
34
|
+
使用 cli 工具,產生 spec 的模板。將你的完整計劃填入到模板之中,並通過 cli 工具生成完整的 architecture diff 讓用戶審閱本次spec的架構設計。
|
|
35
|
+
如果該 spec 設計超過三個模塊,則需要創建 batch spec。
|
|
39
36
|
|
|
40
37
|
## 範例
|
|
41
38
|
- "製作一個網頁德州撲克小遊戲" -> "拆分成多個模塊:遊戲本體邏輯、前端頁面渲染、前端頁面交互邏輯;制定單元測試、整合測試等策略,並製作一份單一的spec指導實作工作。"
|
|
42
|
-
- "提升現有系統的性能" -> "識別目前repo之中拖累性能的代碼。製作batch spec文檔,將repo的全量優化拆分為以三個模塊為一組的優化。對於必須改動業務邏輯才可以做到的性能提升,填寫clarification questions,並等待用戶回答之後更新spec。"
|
|
39
|
+
- "提升現有系統的性能" -> "識別目前 repo 之中拖累性能的代碼。製作 batch spec 文檔,將 repo 的全量優化拆分為以三個模塊為一組的優化。對於必須改動業務邏輯才可以做到的性能提升,填寫 clarification questions,並等待用戶回答之後更新 spec。"
|
|
43
40
|
|
|
44
41
|
## 參考資料
|
|
45
42
|
- `scripts/create-specs` - `apltk create-specs` 背後使用的模板產生器。
|
|
Binary file
|
|
@@ -1,42 +1,40 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: init-project-html
|
|
3
3
|
description: >-
|
|
4
|
-
使用 `apltk architecture` 初始化專案 HTML 架構圖譜,生成基礎 atlas YAML 與渲染後的 HTML
|
|
4
|
+
使用 `apltk architecture` 初始化專案 HTML 架構圖譜,生成基礎 atlas YAML 與渲染後的 HTML 頁面。所有宣告基於倉庫證據;每個功能模組由一個可寫子 agent 負責,主 agent 必須等待全部子 agent 完成後,才能補跨功能連接並執行 render 與 validate。
|
|
5
5
|
---
|
|
6
6
|
|
|
7
|
-
# 初始化專案 HTML 架構圖
|
|
8
|
-
|
|
9
7
|
## 技能目標
|
|
10
8
|
|
|
11
|
-
透過 `apltk architecture` 為目前倉庫建立基礎專案架構圖譜,產出受 CLI 管理的
|
|
9
|
+
透過 `apltk architecture` 為目前倉庫建立基礎專案架構圖譜,產出受 CLI 管理的 atlas 狀態檔與渲染後的 HTML 頁面。
|
|
12
10
|
|
|
13
11
|
## 驗收條件
|
|
14
12
|
|
|
15
13
|
- 只透過 `apltk architecture` 修改圖譜;不得手改 `resources/project-architecture/**/*.html`。
|
|
16
|
-
-
|
|
17
|
-
-
|
|
18
|
-
-
|
|
19
|
-
- 採用「每個功能一個可寫子 agent
|
|
14
|
+
- 所有宣告可追溯到真實程式碼、設定、SQL 或外部邊界;無法確認的部分用 `TBD` 或在 `meta.summary` 記錄遺漏原因。
|
|
15
|
+
- 宏觀圖完整表達功能與子模組關係;所有跨邊界互動使用 `call`、`return`、`data-row`、`failure` 邊表達。
|
|
16
|
+
- 每個非平凡子模組具備足夠內部結構:已宣告 `function`、`variable`、有序 `dataflow`、必要時的 `error`,且引用可通過校驗。
|
|
17
|
+
- 採用「每個功能一個可寫子 agent」分工;主 agent 等所有子 agent 完成後才補跨功能邊,且 `apltk architecture validate` 通過。
|
|
20
18
|
|
|
21
19
|
## 工作流程
|
|
22
20
|
|
|
23
|
-
1.
|
|
24
|
-
2.
|
|
25
|
-
3.
|
|
26
|
-
4.
|
|
27
|
-
5.
|
|
28
|
-
6.
|
|
21
|
+
1. 執行 `apltk architecture --help`,以 CLI 說明為唯一命令真源。
|
|
22
|
+
2. 做整倉淺層盤點,列出功能模組的 slug、使用者視角職責、入口點與主要邊界資源。
|
|
23
|
+
3. 為每個功能派發一個可寫子 agent,負責宣告該功能的全部子模組、函式、變數、資料流、本地錯誤與功能內邊。子 agent 返回子模組摘要及需要主 agent 補上的跨功能邊界資訊。
|
|
24
|
+
4. 主 agent 不得重讀已委派功能的原始碼,也不得重複宣告子 agent 已處理的功能內元件。
|
|
25
|
+
5. 全部子 agent 完成後,主 agent 統一補齊跨功能 edge、必要時補共享 meta 或 actor,然後執行 `apltk architecture render` 與 `apltk architecture validate`。
|
|
26
|
+
6. 抽查渲染結果,確認宏觀圖和至少一個代表性子模組頁滿足驗收條件,彙報功能數、子模組數、邊數量、未覆蓋路徑與原因。
|
|
29
27
|
|
|
30
28
|
## 使用範例
|
|
31
29
|
|
|
32
|
-
- 「替這個倉庫首次生成 HTML
|
|
33
|
-
-
|
|
30
|
+
- 「替這個倉庫首次生成 HTML 架構圖」→ 梳理功能模組,按功能分派子 agent,彙總跨功能邊,生成基礎 atlas 與渲染頁面。交付物為 `resources/project-architecture/` 下的完整 atlas 狀態檔與通過驗證的 HTML 頁面。
|
|
31
|
+
- 「把系統的資料流、呼叫關係和回滾路徑視覺化」→ 使用 `call` / `return` / `data-row` / `failure` 邊表達跨邊界關係,為每個關鍵子模組補齊內部 `dataflow`。
|
|
34
32
|
|
|
35
33
|
## 參考資料索引
|
|
36
34
|
|
|
37
35
|
- `references/TEMPLATE_SPEC.md`:atlas 欄位、列舉和 CLI 寫入形狀速查表。
|
|
38
36
|
- `lib/atlas/cli.js`:`apltk architecture` 的實作入口。
|
|
39
37
|
- `lib/atlas/schema.js`:圖譜資料結構與校驗規則。
|
|
40
|
-
- `sample-demo
|
|
38
|
+
- `sample-demo/`:完整示例輸出,用於理解基礎 atlas 的最終形態。
|
|
41
39
|
- `spec-to-project-html/SKILL.md`:面向規劃文件的 overlay 版本。
|
|
42
40
|
- `update-project-html/SKILL.md`:面向已存在基礎 atlas 的增量刷新版本。
|
|
@@ -19,7 +19,7 @@ description: >-
|
|
|
19
19
|
|
|
20
20
|
- Required: `align-project-documents` and `maintain-project-constraints` after the repository is truly iteration-complete.
|
|
21
21
|
- Conditional: `systematic-debug` when a performance test, benchmark, load run, or production trace exposes a real business-logic defect that must be fixed at the true owner; `improve-observability` when profiling signals are missing or measurement requires durable logs, metrics, or traces.
|
|
22
|
-
- Optional: `
|
|
22
|
+
- Optional: `improve-observability` for profiling signals or durable metrics needed during optimization.
|
|
23
23
|
- Fallback: If required completion dependencies are unavailable, finish performance work and validation first, then report exactly which documentation, constraint-sync, or observability action could not run.
|
|
24
24
|
|
|
25
25
|
## Standards
|
|
@@ -18,7 +18,7 @@ description: >-
|
|
|
18
18
|
|
|
19
19
|
- Required: `align-project-documents` and `maintain-project-constraints` after the repository is truly iteration-complete.
|
|
20
20
|
- Conditional: `systematic-debug` when a newly added or existing test exposes a real business-logic defect that must be fixed at the true owner.
|
|
21
|
-
- Optional: `
|
|
21
|
+
- Optional: `improve-observability` for non-trivial telemetry design.
|
|
22
22
|
- Fallback: If required completion dependencies are unavailable, finish code and validation first, then report exactly which documentation or constraint-sync action could not run.
|
|
23
23
|
|
|
24
24
|
## Standards
|
|
Binary file
|
|
@@ -1,39 +1,35 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: maintain-project-constraints
|
|
3
3
|
description: >-
|
|
4
|
-
依據倉庫現況刷新根目錄 `AGENTS.md` / `CLAUDE.md`,只保留
|
|
5
|
-
`Common Development Commands`、`Project Business Goals`、
|
|
6
|
-
`Project Documentation Index` 三個可追溯區塊。
|
|
4
|
+
依據倉庫現況刷新根目錄 `AGENTS.md` / `CLAUDE.md`,只保留 Common Development Commands、Project Business Goals、Project Documentation Index 三個可追溯區塊。
|
|
7
5
|
---
|
|
8
6
|
|
|
9
|
-
# 維護專案約束文件
|
|
10
|
-
|
|
11
7
|
## 技能目標
|
|
12
8
|
|
|
13
|
-
基於當前倉庫證據,產出或刷新根目錄 `AGENTS.md` 與 `CLAUDE.md
|
|
9
|
+
基於當前倉庫證據,產出或刷新根目錄 `AGENTS.md` 與 `CLAUDE.md`,使其準確反映開發命令、專案商業目標與文件索引,不捏造任何內容。
|
|
14
10
|
|
|
15
|
-
##
|
|
11
|
+
## 驗收條件
|
|
16
12
|
|
|
17
|
-
-
|
|
18
|
-
- `Common Development Commands`
|
|
19
|
-
- `Project Business Goals`
|
|
20
|
-
- `Project Documentation Index`
|
|
13
|
+
- 最終文件正文只包含三個區塊,順序固定:`Common Development Commands` → `Project Business Goals` → `Project Documentation Index`。
|
|
14
|
+
- `Common Development Commands` 中每條命令可追溯到 `package.json`、`Makefile`、`bin/`、`scripts/` 或 CI 設定等真實入口。
|
|
15
|
+
- `Project Business Goals` 只描述專案層級目的、服務對象與交付結果,不展開為功能清單。
|
|
16
|
+
- `Project Documentation Index` 覆蓋現存 `docs/features/`、`docs/architecture/`、`docs/principles/` 與重要根目錄文件,每項對應真實路徑。
|
|
21
17
|
- 過時路徑、虛構命令與多餘區塊已被移除。
|
|
22
|
-
- 若 `AGENTS.md` 與 `CLAUDE.md`
|
|
18
|
+
- 若 `AGENTS.md` 與 `CLAUDE.md` 同時存在且未聲明故意分歧,兩者三個區塊內容一致。
|
|
23
19
|
|
|
24
|
-
##
|
|
20
|
+
## 工作流程
|
|
25
21
|
|
|
26
|
-
1.
|
|
27
|
-
2.
|
|
28
|
-
3. 按專案慣例更新或補齊 `AGENTS.md` / `CLAUDE.md
|
|
22
|
+
1. 從倉庫現況收集可驗證的命令入口、專案目的與現有文件清單,不以舊約束文件作為唯一真相。
|
|
23
|
+
2. 根據證據生成三個必需區塊,確保每條命令、商業目標與文件索引可被直接追溯。
|
|
24
|
+
3. 按專案慣例更新或補齊 `AGENTS.md` / `CLAUDE.md`,正文限制在三個規定區塊內。
|
|
29
25
|
4. 完成前逐項校驗命令來源、文件路徑與雙文件一致性,清除所有陳舊或多餘內容。
|
|
30
26
|
|
|
31
|
-
##
|
|
27
|
+
## 使用範例
|
|
32
28
|
|
|
33
|
-
-
|
|
34
|
-
-
|
|
35
|
-
-
|
|
29
|
+
- 「重建 `docs/` 後,請同步刷新 `AGENTS.md` 和 `CLAUDE.md`」→ 根據當前腳本、入口與文件樹重寫兩個根目錄約束文件,只保留三個規定區塊並確保索引完整。
|
|
30
|
+
- 「這個專案的 `CLAUDE.md` 已經過時,請補一份對應的 `AGENTS.md`」→ 依據現有命令與文件結構建立缺失文件,讓兩份約束文件在預期一致時保持同構。
|
|
31
|
+
- 「請清理根目錄約束文件裡的舊命令與失效路徑」→ 驗證每條命令與索引路徑是否仍成立,只保留能被倉庫證據支持的內容。
|
|
36
32
|
|
|
37
|
-
##
|
|
33
|
+
## 參考資料索引
|
|
38
34
|
|
|
39
|
-
- `references/constraint-file-reference.md
|
|
35
|
+
- `references/constraint-file-reference.md`:三區塊契約、撰寫規則、核對清單與輸出模板。
|
|
@@ -1,51 +1,40 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: merge-changes-from-local-branches
|
|
3
3
|
description: >-
|
|
4
|
-
|
|
5
|
-
安全清理與最終提交準備;除非本次對話中有明確要求,否則不推送遠端。
|
|
4
|
+
將使用者指定的本地分支合併回當前分支,依序完成衝突處理、驗證、安全清理與最終提交準備;除非本次對話明確要求,否則不推送遠端。
|
|
6
5
|
---
|
|
7
6
|
|
|
8
|
-
##
|
|
9
|
-
在工作開始時的當前本地分支上,整合使用者指定的本地分支,或由現有 spec / batch
|
|
10
|
-
規劃可無歧義對應出的本地分支,完成合併、驗證、清理與提交流程銜接,並確保整合結果
|
|
11
|
-
可安全交由 `commit-and-push` 收尾。
|
|
7
|
+
## 技能目標
|
|
12
8
|
|
|
13
|
-
|
|
14
|
-
- 所有實際合併的變更都落在流程開始時的原始當前分支,且不會未經使用者允許改換目標分支。
|
|
15
|
-
- 合併範圍只包含使用者明確指定,或可由 `coordination.md` / spec 上下文無歧義映射出的分支;若映射不清楚,流程會停止並回報。
|
|
16
|
-
- 當 batch 規劃存在 `Merge order` / landing order 時,實際整合順序與規劃一致;若順序衝突或不明確,不會猜測執行。
|
|
17
|
-
- 所有衝突都以保留正確行為為原則完成解決,並在刪除來源分支或交給 `commit-and-push` 前完成必要驗證。
|
|
18
|
-
- 只清理已成功合併且已驗證的來源分支或 worktree;不會強制刪除尚未真正合入目標分支的來源。
|
|
19
|
-
- 最終交付是原始當前分支上的整合結果與簡潔摘要,並只在使用者於本次對話明確要求時才推送遠端;本技能不單獨執行 `archive-specs`。
|
|
20
|
-
|
|
21
|
-
## 工作流程
|
|
22
|
-
1. 確認目標分支與合併範圍
|
|
23
|
-
將流程開始時的當前分支視為唯一權威目標分支。優先使用使用者明確提供的分支名稱建立合併集合;只有在 spec 或 batch `coordination.md` 能無歧義對應時,才可從規劃文件補足分支映射與整合順序。
|
|
9
|
+
在工作開始時的當前本地分支上,整合使用者指定的本地分支(或由 spec / batch 規劃可無歧義對應出的分支),完成合併、驗證、清理,並讓整合結果可安全交由 `commit-and-push` 收尾。
|
|
24
10
|
|
|
25
|
-
|
|
26
|
-
在原始目標分支上確認工作樹適合執行合併。若存在會干擾整合的未提交變更,停止並回報;不要自行 stash、丟棄變更,或切換到其他目標分支。
|
|
11
|
+
## 驗收條件
|
|
27
12
|
|
|
28
|
-
|
|
29
|
-
|
|
13
|
+
- 所有合併變更落在流程開始時的原始當前分支,不會未經允許改換目標分支。
|
|
14
|
+
- 合併範圍只包含使用者明確指定,或可由 `coordination.md` / spec 無歧義映射出的分支;若映射不清楚則停止並回報。
|
|
15
|
+
- 當 batch 規劃存在 `Merge order` / landing order 時,實際整合順序與規劃一致;順序衝突或不明確時不猜測執行。
|
|
16
|
+
- 所有衝突以保留正確行為為原則解決,並在刪除來源分支或交給 `commit-and-push` 前完成驗證。
|
|
17
|
+
- 只清理已成功合併且已驗證的來源分支或 worktree;不強制刪除尚未真正合入的來源。
|
|
18
|
+
- 最終交付是原始當前分支上的整合結果與簡潔摘要;只有使用者於本次對話明確要求時才推送遠端。本技能不單獨執行 `archive-specs`。
|
|
30
19
|
|
|
31
|
-
|
|
32
|
-
先對衝突區域或高風險改動執行對應驗證,再對整體整合結果執行倉庫慣用的標準驗證。任何驗證失敗都必須先在當前分支修正,之後才能進入清理或提交階段。
|
|
20
|
+
## 工作流程
|
|
33
21
|
|
|
34
|
-
|
|
35
|
-
|
|
22
|
+
1. 以流程開始時的當前分支為唯一目標分支,確認合併範圍(使用者明確指定的分支,或由 spec / `coordination.md` 無歧義映射出的分支與整合順序)。
|
|
23
|
+
2. 確認工作樹適合執行合併:若存在干擾整合的未提交變更,停止並回報,不自行 stash 或切換分支。
|
|
24
|
+
3. 依既定順序逐一合併 in-scope 分支。對已合入或無新增內容的分支,跳過並記錄原因。發生衝突時閱讀雙方內容編輯出正確結果;無法在不猜測意圖的前提下解決時停止並回報。必要時使用 `merge-conflict-resolver`。
|
|
25
|
+
4. 先對衝突區域或高風險改動執行針對性驗證,再對整體整合結果執行倉庫慣用的標準驗證。驗證失敗先在當前分支修正。
|
|
26
|
+
5. 僅清理已完成合併且通過驗證的來源分支與對應 worktree;安全刪除被拒絕時保留來源並回報,不使用強制刪除。
|
|
27
|
+
6. 交由 `commit-and-push` 完成必要審查、submission-readiness gate 與本地提交。若 `commit-and-push` 不可用則停止並回報,不走裸 `git commit`。
|
|
28
|
+
7. 總結已合併與跳過的分支、順序依據、衝突處理原則、驗證結果,以及流程最終停在本地 `HEAD` 還是包含遠端推送。
|
|
36
29
|
|
|
37
|
-
|
|
38
|
-
整理合併後的最終提交意圖,並交由 `commit-and-push` 執行必要審查、submission-readiness gate 與最終本地提交。若 `commit-and-push` 不可用,必須停止並回報,不可改走裸 `git commit`。本技能不負責 push、tag、release,也不另外觸發 `archive-specs`;只有使用者在本次對話中明確要求遠端更新時,才允許後續推送。
|
|
30
|
+
## 使用範例
|
|
39
31
|
|
|
40
|
-
|
|
41
|
-
|
|
32
|
+
- 「把 `feature/api-layer` 和 `feature/cli-wrapper` 合回目前分支」→ 以目前分支為唯一目標,依指定順序完成整合、驗證與安全清理,再交給 `commit-and-push` 做本地提交。
|
|
33
|
+
- 「根據 batch 的 `coordination.md` 把各 worktree 分支合回來,但先不要 push」→ 從 batch 規劃確認無歧義分支映射與 landing order,依序合併、驗證、清理,只做到本地提交。
|
|
34
|
+
- 「把那幾個應該相關的分支一起合一下」→ 若無法從使用者輸入或規劃文件明確判定分支集合與順序,停止並回報需要補充資訊。
|
|
42
35
|
|
|
43
|
-
##
|
|
44
|
-
- 「把 `feature/api-layer` 和 `feature/cli-wrapper` 合回目前分支」 -> 以目前分支為唯一目標,依指定順序完成整合、驗證與安全清理,再交給 `commit-and-push` 做最終本地提交。
|
|
45
|
-
- 「根據這個 batch 的 `coordination.md` 把各 worktree 分支合回來,但先不要 push」 -> 先從 batch 規劃確認無歧義分支映射與 landing order,再依序合併、驗證、清理,最後只做到本地提交。
|
|
46
|
-
- 「把那幾個應該相關的分支一起合一下」 -> 若無法從使用者輸入或規劃文件明確判定分支集合與順序,停止並回報需要補充資訊,而不是依 ancestry 或名稱猜測。
|
|
36
|
+
## 參考資料索引
|
|
47
37
|
|
|
48
|
-
## 參考資料
|
|
49
38
|
- `commit-and-push/SKILL.md`:最終提交、submission-readiness 與是否允許推送的權威流程。
|
|
50
39
|
- `merge-conflict-resolver/SKILL.md`:衝突需要精確合成時的輔助技能。
|
|
51
40
|
- `docs/plans/**/coordination.md`:batch 規劃存在時的 landing order 與分支映射依據。
|
|
Binary file
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: open-source-pr-workflow
|
|
3
|
-
description: PR-focused workflow for open-source repositories. Use when the user asks to prepare a PR branch from existing changes, draft/open/update a PR, or push a ready contribution branch. Do not use this skill for implementing product features or editing business logic; use it after code changes are already prepared. Enforce branch naming as codex/{change_type}/{changes}, open PRs directly by default without waiting for draft confirmation, show drafts only when the user explicitly asks to review them first, default forked repositories to open PRs against the upstream parent repository unless the user explicitly requests the fork, and write PR content in English by default with required sections for related issues or motivation, engineering decisions with rationale, and test results with commands. For code-affecting changes, run
|
|
3
|
+
description: PR-focused workflow for open-source repositories. Use when the user asks to prepare a PR branch from existing changes, draft/open/update a PR, or push a ready contribution branch. Do not use this skill for implementing product features or editing business logic; use it after code changes are already prepared. Enforce branch naming as codex/{change_type}/{changes}, open PRs directly by default without waiting for draft confirmation, show drafts only when the user explicitly asks to review them first, default forked repositories to open PRs against the upstream parent repository unless the user explicitly requests the fork, and write PR content in English by default with required sections for related issues or motivation, engineering decisions with rationale, and test results with commands. For code-affecting changes, run code-simplifier before opening the PR.
|
|
4
4
|
---
|
|
5
5
|
|
|
6
6
|
# Open Source PR Workflow
|
|
@@ -8,7 +8,7 @@ description: PR-focused workflow for open-source repositories. Use when the user
|
|
|
8
8
|
## Dependencies
|
|
9
9
|
|
|
10
10
|
- Required: **`commit-and-push`** before publishing the contribution branch (any remaining changes **MUST** be committed/readiness-checked through that skill; **push** only when the user requested remote update—then continue to PR steps).
|
|
11
|
-
- Conditional: `
|
|
11
|
+
- Conditional: `code-simplifier` for code-affecting PRs before opening the PR.
|
|
12
12
|
- Optional: none.
|
|
13
13
|
- Fallback: If **`commit-and-push`** or a **required** code-affecting dependency is unavailable, stop and report the missing dependency instead of bypassing the quality gate.
|
|
14
14
|
|
|
@@ -54,10 +54,7 @@ git checkout -b codex/fix/add-rate-limit-retry
|
|
|
54
54
|
|
|
55
55
|
### 4) Run dependent skills for code-affecting changes
|
|
56
56
|
|
|
57
|
-
- If the PR includes code changes, run `
|
|
58
|
-
- Resolve any confirmed findings before continuing.
|
|
59
|
-
- Then run `code-simplifier` on the updated changes.
|
|
60
|
-
- Keep both passes minimal and focused on the current PR scope.
|
|
57
|
+
- If the PR includes code changes, run `code-simplifier` before opening the PR.
|
|
61
58
|
- Use these skills as internal quality gates only; do not include skill/tool names in PR content.
|
|
62
59
|
- If the PR has no code changes, explicitly note that these two skills were not required in internal notes only.
|
|
63
60
|
|
|
@@ -118,7 +115,7 @@ If tests cannot run locally, state why and provide the closest available validat
|
|
|
118
115
|
- If the user asked to review the draft, they confirmed the final draft before PR creation.
|
|
119
116
|
- If the repository is a fork, PR destination is the upstream parent unless the user explicitly requested the fork.
|
|
120
117
|
- PR body includes all required sections and focuses only on PR-related context.
|
|
121
|
-
- If code changes exist, run `
|
|
118
|
+
- If code changes exist, run `code-simplifier` before opening the PR.
|
|
122
119
|
- PR body does not mention internal skills/tools or agent workflow notes.
|
|
123
120
|
- Test commands and results are explicitly listed.
|
|
124
121
|
- Language defaults to English unless user requests otherwise.
|
package/optimise-skill/SKILL.md
CHANGED
|
@@ -12,19 +12,19 @@ description: Use prompt engineering to optimise agent skill. Use it when you nee
|
|
|
12
12
|
|
|
13
13
|
## 工作流程
|
|
14
14
|
|
|
15
|
-
1. 識別交付物
|
|
15
|
+
### 1. 識別交付物
|
|
16
16
|
完整閱讀整個技能的 `SKILL.md` 文檔,以及其引用的所有額檔案,包括但不限於 `references/` , `scripts/` 。基於獲取到的技能上下文資訊,總結出該技能的最終交付產物。
|
|
17
17
|
|
|
18
|
-
2. 制定驗收條件
|
|
18
|
+
### 2. 制定驗收條件
|
|
19
19
|
制定驗收條件,從而確保LLM能夠穩定、可復現地按照驗收條件產出符合要求高質量最終交付物
|
|
20
20
|
|
|
21
|
-
3. 重寫技能
|
|
21
|
+
### 3. 重寫技能
|
|
22
22
|
將整個技能的 `SKILL.md` 重寫。重寫後的技能應該只包含以下幾個關鍵組成部分:
|
|
23
|
-
-
|
|
24
|
-
-
|
|
25
|
-
-
|
|
26
|
-
-
|
|
27
|
-
-
|
|
23
|
+
- 目標
|
|
24
|
+
- 驗收條件
|
|
25
|
+
- 工作流程
|
|
26
|
+
- 範例
|
|
27
|
+
- 參考資料
|
|
28
28
|
對於技能有幫助的內但不符合上述技能架構規範,則應該被分類放置在 `references/` 下的markdown檔案。
|
|
29
29
|
|
|
30
30
|
## 範例
|
|
@@ -12,19 +12,19 @@ description: Use prompt engineering to optimise agent skill. Use it when you nee
|
|
|
12
12
|
|
|
13
13
|
## 工作流程
|
|
14
14
|
|
|
15
|
-
1. 識別交付物
|
|
15
|
+
### 1. 識別交付物
|
|
16
16
|
完整閱讀整個技能的 `SKILL.md` 文檔,以及其引用的所有額檔案,包括但不限於 `references/` , `scripts/` 。基於獲取到的技能上下文資訊,總結出該技能的最終交付產物。
|
|
17
17
|
|
|
18
|
-
2. 制定驗收條件
|
|
18
|
+
### 2. 制定驗收條件
|
|
19
19
|
制定驗收條件,從而確保LLM能夠穩定、可復現地按照驗收條件產出符合要求高質量最終交付物
|
|
20
20
|
|
|
21
|
-
3. 重寫技能
|
|
21
|
+
### 3. 重寫技能
|
|
22
22
|
將整個技能的 `SKILL.md` 重寫。重寫後的技能應該只包含以下幾個關鍵組成部分:
|
|
23
|
-
-
|
|
24
|
-
-
|
|
25
|
-
-
|
|
26
|
-
-
|
|
27
|
-
-
|
|
23
|
+
- 目標
|
|
24
|
+
- 驗收條件
|
|
25
|
+
- 工作流程
|
|
26
|
+
- 範例
|
|
27
|
+
- 參考資料
|
|
28
28
|
對於技能有幫助的內但不符合上述技能架構規範,則應該被分類放置在 `references/` 下的markdown檔案。
|
|
29
29
|
|
|
30
30
|
## 範例
|
package/package.json
CHANGED
|
Binary file
|
|
Binary file
|