Skip to content

Releases: franklinxkk/ai-delivery-spec

v4.9.14

Choose a tag to compare

@franklinxkk franklinxkk released this 04 Jul 14:43

1.add oa-domain knowledge
2.harden validation and domain routing
3.fixed some bug

v4.9.10 - Multi-Module PRD Quality Gates

Choose a tag to compare

@franklinxkk franklinxkk released this 30 Jun 16:31

v4.9.10 - 2026-06-30

  • Added Stage 3.5 Cross-Module Flow Contract for multi-module scopes.
  • Added Post-Generation Multi-Module Checklist to block PASS when split PRDs lose flow closure, FRR completeness, state/event mapping, prototype anchors, or domain gates.
  • Added Domain Knowledge Quality Gate and First-Principles Domain Lens across domain modules.
  • Added CRM-specific gates for cross-module handoff, R01-R12 alert back-references, ER/field/event/notification integrity, and prototype testability.
  • Kept compact skill structure: no new reference files added.

v4.9.7 - agentskill.sh ASCII compatibility

Choose a tag to compare

@franklinxkk franklinxkk released this 27 Jun 13:46

What's New in v4.9.7

  • Made root SKILL.md ASCII-only for agentskill.sh compatibility.
  • Replaced the remaining em dash and multiplication-sign characters in the runtime file.
  • Kept frontmatter in the simplest name and single-line description form.

Validation

  • validate_skill_consistency.py
  • validate_routing_scenarios.py
  • validate_release_readiness.py --crm-dir D:\work\codex\crm

Full changelog: https://github.com/franklinxkk/ai-delivery-spec/blob/main/CHANGELOG.md

v4.9.6 - agentskill.sh parser compatibility

Choose a tag to compare

@franklinxkk franklinxkk released this 27 Jun 13:27

What's New in v4.9.6

  • Further simplified SKILL.md frontmatter for agentskill.sh compatibility.
  • Uses a short single-line description with no folded YAML, no colon-space sequence, and no slash-heavy parser edge cases.
  • Preserves practical exclusions for code debugging and copy rewriting.
  • Synchronized version metadata and validators.

Validation

  • validate_skill_consistency.py
  • validate_routing_scenarios.py
  • validate_release_readiness.py --crm-dir D:\work\codex\crm

Full changelog: https://github.com/franklinxkk/ai-delivery-spec/blob/main/CHANGELOG.md

v4.9.5 - agentskill.sh frontmatter compatibility

Choose a tag to compare

@franklinxkk franklinxkk released this 27 Jun 12:53

What's New in v4.9.5

  • Fixed agentskill.sh import compatibility by changing SKILL.md YAML frontmatter description from folded block syntax to a standard single-line string.
  • Preserved the trigger exclusions for code-only syntax/debugging, copy rewriting, and idea exploration with no delivery intent.
  • Synchronized version metadata across README, SKILL.md, OpenAI metadata, templates, community submission docs, and CHANGELOG.

Validation

  • validate_skill_consistency.py
  • validate_routing_scenarios.py
  • validate_release_readiness.py --crm-dir D:\work\codex\crm

Full changelog: https://github.com/franklinxkk/ai-delivery-spec/blob/main/CHANGELOG.md

v4.9.4 - Community README and release readiness

Choose a tag to compare

@franklinxkk franklinxkk released this 27 Jun 11:49

What's New in v4.9.4

  • Reworked README for GitHub/community discovery and first-screen conversion.
  • Added a clearer pain statement for PMs, developers, QA, AI-native teams, and coding-agent users.
  • Added direct quick-start prompts for rough ideas, PRD review, Human-First PRD, AI-Coding PRD, and prototype/competitor reverse engineering.
  • Clarified work paths, PRD profiles, install paths, delivery package convention, toolchain integration, and domain modules.

Included From v4.9.3

  • skills.sh badge and npx skills add franklinxkk/ai-delivery-spec install path.
  • Validator hardening for PRD quality, bilingual routing, domain-module sync, and release readiness simulation.
  • Release readiness script covering traditional PRD, AI-native, AI-coding, PRD review, competitor/prototype, lifecycle, mobile, launch, and retirement paths.

Validation

  • validate_skill_consistency.py
  • validate_routing_scenarios.py
  • validate_release_readiness.py --crm-dir D:\work\codex\crm

Full changelog: https://github.com/franklinxkk/ai-delivery-spec/blob/main/CHANGELOG.md

v4.6.2 — Medical/Hospital IT Domain + Coding-Agent Compatibility

Choose a tag to compare

@franklinxkk franklinxkk released this 21 Jun 00:06

What's New in v4.6.2

Hospital IT Domain Module

  • references/domain-medical-hospital-it.md — HIS/EMR/LIS/PACS/RIS, clinical workflow, medical quality, infection control, internet hospital, research data governance, AI-assisted medical operations

Coding-Agent Compatibility Hardened

  • AI contract selection ladder
  • AC ID evolution rules
  • OpenAI/Codex coding-agent entrypoint
  • Deterministic PRD/prototype data-* mapping validator

New Example

  • examples/medical-hospital-it/README.md

What's Included

4 Entry Points: SKILL.md + delivery-core.md + prototype-testability.md + advanced-extensions.md

4 Domain Modules: Traffic Safety, CRM, Higher-Education IT, Medical/Hospital IT

Coding-Agent Compatibility: agents/claude-code.md + agents/openai-codex.md + agents/openai.yaml + coding-agent-compat.md + 5 validation scripts

Quick Start

Use AI Delivery Spec. Run 0D triage: [TIER] [AI] [WORKFLOW]. Load only relevant entrypoints.

Complementary with github/spec-kit

spec-kit handles spec-to-code, ai-delivery-spec handles requirement-to-spec+prototype.

Full Changelog: https://github.com/franklinxkk/ai-delivery-spec/blob/main/CHANGELOG.md

v4.5.2 — Higher-Education Domain + PRD Runtime Consistency

Choose a tag to compare

@franklinxkk franklinxkk released this 18 Jun 15:53

What's New in v4.5.2

🎓 New Domain Module: Higher-Education Informationization

eferences/domain-education-it.md — academic affairs, student affairs, teaching quality, smart classroom, one-stop service, data governance, campus AI assistant

🔧 PRD Runtime Consistency Fix

  • FRR summary now has the same 16 sections as the standard PRD template
  • Function-Level NFR separated from dependencies
  • PRD readability explicitly covers Sponsor/PM, UX/UI, Frontend, Backend, Algorithm/AI, and QA readers

📋 What's Included

4 Entry Points:

  • SKILL.md — triage, routing, gates

eferences/delivery-core.md — PRD, stories, DDD/API/data, lifecycle

eferences/prototype-testability.md — prototype, mobile, interaction

eferences/advanced-extensions.md — AI, SaaS, approval, reporting, global

3 Domain Modules:

  • Traffic Safety / 交通安全
  • CRM / 客户经营
  • Higher-Education IT / 高校教育信息化

Validation Scripts:

  • �alidate_skill_consistency.py
  • �alidate_routing_scenarios.py

💡 Quick Start

Use AI Delivery Spec as the delivery standard. First run 0D triage: [TIER] [AI] [WORKFLOW]. Load only the relevant entrypoint files. Produce the requested artifact and end with gates, verification, gaps, and completion state.

🔄 Complementary with github/spec-kit

spec-kit handles spec→code, ai-delivery-spec handles requirement→spec+prototype.


Full Changelog: https://github.com/franklinxkk/ai-delivery-spec/commits/main