ai-developer-skill-os 8.1.5 → 8.1.7

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.
Files changed (4) hide show
  1. package/.agents/CHANGELOG.md +113 -70
  2. package/.agents/README.md +44 -58
  3. package/CHANGELOG.md +22 -0
  4. package/package.json +1 -1
package/.agents/CHANGELOG.md CHANGED
@@ -1,70 +1,113 @@
1
- # Changelog
2
-
3
- All notable changes to the **AI Developer Skill OS** project will be documented in this file.
4
-
5
- The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
6
- and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
7
-
8
- ## [7.0.2] - 2026-07-15
9
-
10
- ### Added
11
- - **Frontend Integration**: Added `qk-fe-api-integration` skill to strictly enforce zero-overhead base API consumption, DTO mapping, and resilient UI states.
12
-
13
- ## [7.0.0] - 2026-07-15
14
-
15
- ### Changed (The "Anti-Slop & Zero-Trust" Architecture Upgrade)
16
- - **Zero-Trust Logic**: Core logic skills (`qk-context-loader`, `qk-orchestrator`, `qk-feature-delivery`) now mandate a Topological Context Graph prior to any code generation.
17
- - **Strict Design Contract**: Frontend skills (`qk-ui-system-builder`, `qk-ui-builder`, `qk-ui-audit`) now explicitly reject generic styling (slop) and enforce strict token matching against a mandated `DESIGN.md`.
18
- - **Repair Loop Enforcement**: `qk-bug-resolution` updated to strictly prohibit jumping to fixes without explicitly establishing an `Observe -> Hypothesis -> Evidence` chain.
19
- - **Contract-First APIs**: `qk-api-lifecycle` now requires explicit OpenAPI/Swagger definitions prior to endpoint implementation.
20
- - **Upgraded Testing Framework**: Fixed regex and validation logic in `behavior-conformance.test.js` to rigorously support YAML array strings and precise V7 markdown constraints. All 22 skills bumped to `v7.0.0`.
21
- - **Project Bootstrapping**: `qk-project-bootstrap` now scaffolds `DESIGN.md` explicitly upon initialization.
22
-
23
-
24
- ## [4.0.0] - 2026-07-02
25
-
26
- ### Changed (Major Architecture Upgrade)
27
- - **Intent-Based Agent Architecture:** Replaced rule-based individual skill exceptions with global behavioral group classifications (Static Analysis, Development, Validation, Maintenance).
28
- - **OS Kernel Refactor:** Completely rewrote `AGENTS.md` (the "OS Kernel") into a strictly organized, lightweight structure under 100 lines.
29
- - **Progressive Evidence Collection:** Agents now enforce an 80% Confidence Threshold and incremental Context Budget (`1 file` → `3 files`) to prevent over-fetching and hallucination loops.
30
- - **Cost & Escalation Policies:** Introduced strict cost-optimization logic, Stopping Criteria, and Escalation protocols to prevent infinite `run_command` retries.
31
- - **Risk-based Verification:** Dynamic test execution based on the risk level of changes (Levels 0-3).
32
-
33
- ### Added
34
- - **Registry Validation Suite:** Full Vitest test suite enforcing SKILL.md frontmatter compliance against `docs/SPEC.md`.
35
- - **GitHub Actions CI:** Automated validation on every PR and push to main.
36
- - **Idempotent Installer:** `bin/install.js` now detects and updates existing rule blocks instead of duplicating them.
37
- - **Kilo Code Native Support:** First-class support for `CLAUDE.md` + `kilo.json` config generation.
38
- - **Governance Docs:** `docs/GOVERNANCE.md`, ADRs, and skill classification policy for enterprise maintainability.
39
- - **Engineering Standards:** `qk-engineering-standard` and `qk-validation-gate` skills enforce SOLID, DRY, Clean Code, and mandatory quality gates.
40
-
41
- ## [1.0.1] - 2026-07-01
42
- ### Fixed
43
- - Standardized Language rule across all 23 SKILL.md files for consistency.
44
- - Updated registry version to match package.json.
45
-
46
- ## [1.0.0] - 2026-07-01
47
-
48
- ### Added
49
- - Released the complete AI Developer Skill OS with 23 targeted skills for coding agents.
50
- - **Engineering Core (8 skills):**
51
- - `agent-orchestrator` for planning and delegating tasks.
52
- - `context-manager` for project context and file selection.
53
- - `project-audit` for 3-mode health checks.
54
- - `bug-fix` with root cause analysis.
55
- - `refactor` for safe, behavior-preserving code restructuring.
56
- - `migration` for dependency updates and library swapping.
57
- - `api-integration` for robust API connections.
58
- - `git-engineer` for commit messages and PR documentation.
59
- - **Frontend Core (11 skills):**
60
- - `frontend-architecture`, `design-system`, `ui-builder`, `component-generator`.
61
- - `state-management`, `form-builder`, `table-crud-generator`.
62
- - `frontend-debug`, `frontend-testing`, `accessibility-audit`, `frontend-performance`.
63
- - **Backend Core (4 skills):**
64
- - `backend-architecture`, `database-engineer`, `auth-security`, `deployment`.
65
- - Added comprehensive `skills.json` registry with dependencies and trigger keywords.
66
- - Added bilingual `README.md` (English/Vietnamese).
67
- - Added detailed user documentation in `docs/HUONG_DAN_SU_DUNG.md` and `docs/CHI_TIET_SKILLS.md`.
68
- - Translated all skill frontmatter descriptions to Vietnamese for better UX in agent terminals.
69
- - Included generic English and Vietnamese usage examples in `_template/examples/`.
70
-
1
+ # Changelog
2
+
3
+ All notable changes to AI Developer Skill OS are documented here.
4
+ Format: [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
5
+
6
+ ---
7
+
8
+ ## [8.1.7] — 2026-07 — README Sync
9
+
10
+ ### Changed
11
+ - **README Sync**: Synced `.agents/README.md` with root `README.md` so the installed documentation is fully up-to-date with V8.1 architecture.
12
+ - Bumped version to `8.1.7`.
13
+
14
+ ---
15
+
16
+ ## [8.1.6] — 2026-07 — Design Intelligence Content & Sync
17
+
18
+ ### Added
19
+ - **Design Intelligence Populated**: Sourced core market-standard knowledge (Tailwind, Radix, Apple, Vercel-like principles) into `.agents/knowledge/design-intelligence/` covering components, patterns, industries, themes, and references.
20
+ - **Animation Knowledge**: Added `visual-language/animation.md` for micro-interactions and transition durations.
21
+ - **Dependency Wiring**: Linked `qk-ui-system-builder` to `design-intelligence` so it utilizes the newly generated design constraints.
22
+ - **Changelog Sync**: Synced `.agents/CHANGELOG.md` with root `CHANGELOG.md` so the installed copy reflects the true V8 evolution.
23
+
24
+ ### Changed
25
+ - All skill descriptions properly translated to Vietnamese for ecosystem consistency.
26
+ - Bumped version to `8.1.6` for final, truly complete Design Intelligence NPM Release.
27
+
28
+ ---
29
+
30
+ ## [8.1.4] — 2026-07 — Design Intelligence Pack
31
+
32
+ ### Added
33
+ - **Design Intelligence Layer**: `.agents/knowledge/design-intelligence/` (industries, themes, patterns, visual-language, components, templates)
34
+ - **Knowledge Dependencies**: `qk-design-system-engineering` and `qk-ui-builder` now have strict dependency bounds tied to Design Intelligence without bloating the Capability Graph.
35
+ - **Design Routing Test Cases**: `design-intel-01`, `02`, `03` for validating routing logic in design scenarios.
36
+ - **Re-introduced CLI Installer**: Restored and refactored `bin/install.js` to correctly copy `.agents` across global/local directories for easy NPM setup.
37
+
38
+ ### Changed
39
+ - Refined **Decision Boundaries**: `qk-ui-builder` explicitly forbidden from defining visual direction or brand identity (strict capability bounds).
40
+ - **README.md** reflects 30 Master Skills and the "Agent Engineering OS" mindset.
41
+ - **package.json**: Version bump `8.1.4` and whitelisted files for npm publish.
42
+
43
+ ---
44
+
45
+ ## [8.1.3] — 2026-07 — Architecture Freeze & Governance
46
+
47
+ ### Added
48
+ - **Capability Graph & Routing**: `.agents/registry/capability-graph.yml` to define skill relations (`depends_on`, `delegates_to`).
49
+ - **Priority Governance**: `.agents/rules/priorities.yml` enforcing safety layers (`P1: System Integrity`, `P2: Guardrails`).
50
+ - **Audit Tools**: `tooling/run-aar.js` for architectural acceptance testing.
51
+ - Test cases for routing and boundary attacks.
52
+
53
+ ### Changed
54
+ - Name transition from "Self-Improving" to **Adaptive Agent OS** (Self Optimizing ≠ Self Modifying).
55
+ - Enforced hard cap of **30 Skills** to prevent Skill Inflation.
56
+ - Frozen Baseline achieved: `v8.1.3-agent-os-production`.
57
+
58
+ ---
59
+
60
+ ## [8.0.0] — 2026-07 — Agent Knowledge System
61
+
62
+ ### Philosophy shift
63
+ - **V7:** Skill Repository — human selects skill, AI executes
64
+ - **V8:** Agent Knowledge System — AI retrieves context, selects capability, executes workflow, verifies
65
+
66
+ ### Added
67
+ - `rules/` — Agent behavior policies (global, coding, safety)
68
+ - `workflows/` — Reusable execution pipelines (YAML, with inputs/outputs per step)
69
+ - `examples/` — Few-shot knowledge (good/, bad/)
70
+ - `learnings/` — Evidence-based learning lifecycle (draft/, validated/, deprecated/)
71
+ - `registry/` — Generated AI retrieval index (skills-index.yml)
72
+ - `docs/ARCHITECTURE.md` — Agent request flow diagram
73
+ - `docs/VERSIONING.md` — Semantic versioning policy
74
+ - `docs/schemas/` — Schema definitions for all knowledge objects
75
+ - `docs/decisions/` — Architecture Decision Records (ADRs)
76
+ - `.agentsignore` — AI boundary file (what agents should NOT load)
77
+ - `tooling/` — Dev automation (generate-registry.js, validate-skills.js)
78
+ - `tests/agent-evaluation/` — Agent behavior test harness
79
+
80
+ ### Changed
81
+ - SKILL.md schema upgraded to V8: added `type`, `intent`, `triggers`, `complexity`, `workflow`, `verification`, `related_skills`
82
+ - `skills.json` demoted to **generated compatibility manifest** (source of truth = SKILL.md)
83
+ - Workflows extracted from individual skills → shared reusable templates
84
+
85
+ ### Removed
86
+ - Embedded workflows inside SKILL.md body (moved to workflows/)
87
+ - Old migration scripts (migrate-v*.js, patch*.py, add_lang.*, etc.)
88
+ - Duplicate documentation (CLAUDE.md, old docs/)
89
+ - Experimental packages (packages/edaos-sdk)
90
+ - framework/, knowledge/, specs/, templates/ directories
91
+
92
+ ### Migration
93
+ - V7.5.1 tagged at git: `v7.5.1`
94
+ - V7 skills preserved in git history; migration is additive schema upgrade
95
+
96
+ ---
97
+
98
+ ## [7.5.1] — 2026-07 — Final V7 Release
99
+
100
+ - Bumped all skill versions to 7.5.1
101
+ - Added qk-fe-api-integration skill
102
+ - Added qk-performance-tuner (global config)
103
+ - Stabilized frontmatter schema across all 22 skills
104
+
105
+ ## [7.5.0] — 2026-07
106
+
107
+ - Major skills refactor with EDAOS runtime adapter
108
+ - Added MCP server integration
109
+ - Added qk-bug-resolution workflow contracts
110
+
111
+ ## [7.0.0] — 2026
112
+
113
+ - V7 release: 22 skills, YAML frontmatter, token budget system
package/.agents/README.md CHANGED
@@ -1,74 +1,52 @@
1
- # 🚀 AI Developer Skill OS (ai-developer-skill-os) v7.5.0
1
+ # 🚀 AI Developer Skill OS (ai-developer-skill-os) v8.1.4
2
2
 
3
- > **V7.5.0 "Anti-Slop & Zero-Trust Architecture"**
3
+ > **v8.1.4 "Agent Engineering OS + Design Intelligence"**
4
4
 
5
- Hệ sinh thái AI Developer Skill OS đã được tái cấu trúc hoàn toàn. Thay vì cung cấp các "công cụ rời rạc" (Toolbox), phiên bản V7 được thiết kế như một **Hệ điều hành khép kín, hoạt động theo kỷ luật của một Kiến trúc sư Hệ thống (50 năm kinh nghiệm)**.
5
+ Hệ sinh thái AI Developer Skill OS đã lột xác hoàn toàn. Từ một bộ "công cụ phân tán" (Toolbox) ở V7, phiên bản V8.1.4 được thiết kế như một **Hệ điều hành khép kín (Agent Engineering OS)**, sở hữu năng lực giám sát, bảo vệ ranh giới quyết định (Decision Boundaries) và được trang bị thêm tầng **Design Intelligence**.
6
6
 
7
7
  ---
8
8
 
9
- ## 🏗️ Triết Lý Hoạt Động Cốt Lõi (V7 Kernel)
9
+ ## 🏗️ Triết Lý Hoạt Động Cốt Lõi (V8 Kernel)
10
10
 
11
- ### 1. Zero-Trust Context (Không bao giờ đoán mò)
12
- Mọi kỹ năng phân tích logic và backend (như `qk-feature-delivery`, `qk-orchestrator`) đều bị khoá chặt, không được phép hoạt động nếu chưa có **Dependency Graph** (Bản đồ cấu trúc) từ `qk-context-loader`. AI không được phép thay đổi code dựa trên trí tưởng tượng.
11
+ ### 1. Decision Boundary (Ranh giới Thẩm quyền)
12
+ Không còn tình trạng một Skill ôm đồm mọi việc. Mọi Skill (Kỹ năng) hiện tại bị quản lý bởi các trường `owns`, `does_not_own`, `conflicts_with`, `delegates_to`. Thợ code UI (`qk-ui-builder`) không được phép tự ra quyết định kiến trúc, và ngược lại.
13
13
 
14
- ### 2. Contract-First (Khế ước đi đầu)
15
- - **Frontend / UI**: Mọi sửa đổi giao diện đều phải dựa trên `DESIGN.md`. Nếu vi phạm (tạo ra giao diện lười biếng "slop", dùng màu mặc định sai lệch), AI sẽ tự động bị đánh trượt bài kiểm toán.
16
- - **Backend / API**: Mọi API phải được viết `OpenAPI/Swagger` schema trước khi viết logic thực tế (`qk-api-lifecycle`).
17
- - **Database**: Cấm truy vấn mù. Bắt buộc phải có Schema định nghĩa rõ ràng trước (`qk-data-lifecycle`).
14
+ ### 2. Design Intelligence Layer
15
+ Khắc phục điểm yếu "mù thẩm mỹ" của AI. Agent giờ đây không chỉ biết code mà còn có Gu Thẩm Mỹ (Design Taste). Dựa trên yêu cầu của bạn, hệ thống tự động tra cứu **Knowledge Graph** để chọn đúng Theme (Fintech, Healthcare, SaaS...), Typography và Visual Patterns trước khi giao việc cho UI Builder.
18
16
 
19
- ### 3. Repair Loop (Chu trình Sửa Lỗi Khép Kín)
20
- Bảo vệ mã nguồn khỏi thói quen "sửa vội" của AI. Lỗi bắt buộc phải qua chu trình điều tra nghiêm ngặt: `Observe -> Hypothesis -> Evidence -> Fix -> Verify` (`qk-bug-resolution`).
17
+ ### 3. Khế Ước Đi Đầu (Contract-First & Zero-Trust)
18
+ - **Frontend / UI**: Mọi sửa đổi phải tuân thủ Design System (được `qk-design-system-engineering` kiểm duyệt).
19
+ - **Backend / API**: Phải viết OpenAPI/Swagger contract trước (`qk-api-lifecycle`).
20
+ - **Data**: Cấm AI tự đoán cấu trúc DB, mọi truy vấn phải có Schema (`qk-data-lifecycle`).
21
21
 
22
22
  ---
23
23
 
24
- ## 🧩 Danh sách 22 Master Skills
24
+ ## 🧩 Danh sách 30 Master Skills
25
25
 
26
- Các kỹ năng được quản lý cực kỳ chặt chẽ với bài kiểm tra tự động (Vitest) để đảm bảo tính nhất quán của luật lệ:
27
- - **Orchestration:** `qk-orchestrator`, `qk-context-loader`, `qk-access-policy`
28
- - **Engineering / Dev:** `qk-feature-delivery`, `qk-api-lifecycle`, `qk-fe-api-integration`, `qk-data-lifecycle`, `qk-ui-builder`, `qk-ui-system-builder`
29
- - **Validation & Standards:** `qk-validation-gate`, `qk-engineering-standard`, `qk-ui-audit`, `qk-project-health`, `qk-bug-resolution`
30
- - **Ops & AI:** `qk-system-evolution`, `qk-production-release`, `qk-ai-builder`, `qk-project-bootstrap`
31
- - **Docs & Utils:** `qk-docs`, `qk-project-memory`, `qk-help`, `qk-db-optimizer`
26
+ Các kỹ năng được quản lý bằng **Capability Graph** và được đánh giá liên tục thông qua Evaluation Suite. Bao gồm 7 lĩnh vực:
27
+ 1. **Core & Orchestration:** `qk-orchestrator`, `qk-context-loader`, `qk-project-memory`, `qk-help`
28
+ 2. **Product & Architecture:** `qk-product-specification`, `qk-project-bootstrap`, `qk-frontend-architecture`
29
+ 3. **Frontend & UI:** `qk-design-system-engineering`, `qk-ui-system-builder`, `qk-ui-builder`, `qk-fe-api-integration`, `qk-ui-audit`
30
+ 4. **Backend & Data:** `qk-api-lifecycle`, `qk-data-lifecycle`, `qk-access-policy`
31
+ 5. **Engineering & Delivery:** `qk-feature-delivery`, `qk-bug-resolution`, `qk-engineering-standard`, `qk-system-evolution`, `qk-docs`
32
+ 6. **Quality & Testing:** `qk-test-engineering`, `qk-validation-gate`, `qk-web-quality-gate`, `qk-project-health`
33
+ 7. **Security, DevOps & Analytics:** `qk-security-audit`, `qk-db-optimizer`, `qk-devops-platform`, `qk-production-release`, `qk-agent-observability`, `qk-ai-builder`
32
34
 
33
35
  ---
34
36
 
35
37
  ## 💻 Cách Cài Đặt (Installation)
36
38
 
37
- ### Cách 1: NPM Global (Khuyên dùng cho nhiều máy)
39
+ ### Cách 1: NXP (Không cần cài global) - MỚI
38
40
  ```bash
39
- npm i -g ai-developer-skill-os
41
+ npx ai-developer-skill-os init
40
42
  ```
41
43
 
42
- ### Cách 2: NXP (Không cần cài global)
44
+ ### Cách 2: NPM Global (Khuyên dùng cho cá nhân)
43
45
  ```bash
44
- npx ai-developer-skill-os init --ide=antigravity --scope=2
45
- ```
46
-
47
- ### Cách 3: NPM Scripts (Trong dự án)
48
- ```bash
49
- # Antigravity Global — 1 lần cài, mọi dự án dùng chung
50
- npm run install:antigravity
51
-
52
- # Antigravity Local — cài riêng cho dự án này
53
- npm run install:antigravity:local
54
-
55
- # Các IDE khác
56
- npm run install:cursor # Global Cursor
57
- npm run install:windsurf # Global Windsurf
58
- npm run install:claude # Global Claude Code
59
- npm run install:kilo # Global Kilo Code
60
- npm run install:multi # Multi-IDE
46
+ npm i -g ai-developer-skill-os
61
47
  ```
62
48
 
63
- ## 📦 Tham số cho `init`
64
-
65
- | Tham số | Giá trị | Mô tả |
66
- |---------|---------|-------|
67
- | `--ide` | `antigravity`, `cursor`, `windsurf`, `claude`, `kilo`, `multi-ide` | IDE/AI Assistant |
68
- | `--scope` | `1` = Local, `2` = Global | Phạm vi cài đặt |
69
-
70
- ### Ví dụ thực tế
71
-
49
+ ### Chạy bằng tham số (Command Line Arguments)
72
50
  ```bash
73
51
  # Cài Antigravity Global (toàn máy)
74
52
  npx ai-developer-skill-os init --ide=antigravity --scope=2
@@ -78,23 +56,31 @@ npx ai-developer-skill-os init --ide=antigravity --scope=1
78
56
 
79
57
  # Cài Cursor Local
80
58
  npx ai-developer-skill-os init --ide=cursor --scope=1
81
-
82
- # Cài Multi-IDE (tạo CLAUDE.md + .qk-ai-skill-os)
83
- npx ai-developer-skill-os init --ide=multi-ide --scope=2
84
59
  ```
85
60
 
86
- ## 🧪 Testing
61
+ ## 📦 Các IDE/AI Assistant Được Hỗ Trợ
62
+ - (1) Cursor
63
+ - (2) Windsurf
64
+ - (3) Cline / Roo Code
65
+ - (4) Antigravity / Gemini
66
+ - (5) Codex
67
+ - (6) Kilo Code
87
68
 
88
- Hệ thống được bảo vệ bằng bộ Test Suite (Vitest) để tự động kiểm toán tính hợp lệ của toàn bộ 22 file Kỹ năng trước khi xuất bản:
69
+ ---
70
+
71
+ ## 🧪 Đánh Giá Tự Động (Agent Evaluation)
72
+
73
+ Hệ thống đi kèm một bộ Runner nội bộ để kiểm toán **Routing Intelligence**. Bất kỳ cập nhật nào vào hệ thống Kỹ năng đều phải vượt qua bài kiểm tra hóc búa (Bypass Security, Boundary Attacks, Ambiguous Prompts):
89
74
 
90
75
  ```bash
91
- npm test
76
+ npm run test:agent
77
+ npm run test:graph
78
+ npm run test:registry
92
79
  ```
93
80
 
94
- ## 🚀 Tra Cứu (Help)
95
-
96
- Để tra cứu danh sách các Kỹ năng và luật lệ khắt khe của hệ thống V7, hãy gọi:
81
+ ## 🚀 Hướng Dẫn Sử Dụng
82
+ Sau khi cài đặt thành công, hãy gõ lệnh:
97
83
  ```bash
98
84
  ./qk-help
99
85
  ```
100
-
86
+ để kích hoạt AI Orchestrator và khám phá sức mạnh của hệ điều hành!
package/CHANGELOG.md CHANGED
@@ -5,6 +5,28 @@ Format: [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
5
5
 
6
6
  ---
7
7
 
8
+ ## [8.1.7] — 2026-07 — README Sync
9
+
10
+ ### Changed
11
+ - **README Sync**: Synced `.agents/README.md` with root `README.md` so the installed documentation is fully up-to-date with V8.1 architecture.
12
+ - Bumped version to `8.1.7`.
13
+
14
+ ---
15
+
16
+ ## [8.1.6] — 2026-07 — Design Intelligence Content & Sync
17
+
18
+ ### Added
19
+ - **Design Intelligence Populated**: Sourced core market-standard knowledge (Tailwind, Radix, Apple, Vercel-like principles) into `.agents/knowledge/design-intelligence/` covering components, patterns, industries, themes, and references.
20
+ - **Animation Knowledge**: Added `visual-language/animation.md` for micro-interactions and transition durations.
21
+ - **Dependency Wiring**: Linked `qk-ui-system-builder` to `design-intelligence` so it utilizes the newly generated design constraints.
22
+ - **Changelog Sync**: Synced `.agents/CHANGELOG.md` with root `CHANGELOG.md` so the installed copy reflects the true V8 evolution.
23
+
24
+ ### Changed
25
+ - All skill descriptions properly translated to Vietnamese for ecosystem consistency.
26
+ - Bumped version to `8.1.6` for final, truly complete Design Intelligence NPM Release.
27
+
28
+ ---
29
+
8
30
  ## [8.1.4] — 2026-07 — Design Intelligence Pack
9
31
 
10
32
  ### Added
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "ai-developer-skill-os",
3
- "version": "8.1.5",
3
+ "version": "8.1.7",
4
4
  "description": "Agent Engineering OS: A deterministic capability graph, governance rules, and dynamic knowledge base for AI coding agents.",
5
5
  "main": "bin/install.js",
6
6
  "files": [