feat: integrate prepare-delivery and gate-and-ship plugins#318
feat: integrate prepare-delivery and gate-and-ship plugins#318
Conversation
Add two new plugins to the agentsys ecosystem: - prepare-delivery: Pre-ship quality gates (deslop, simplify, agnix, enhance, review loop, delivery validation, docs sync) - gate-and-ship: Orchestrator that chains /prepare-delivery then /ship Updated: - marketplace.json: add both plugins (17 -> 19 plugins) - generate-docs.js: add CATEGORY_MAP, STATIC_SKILLS, PURPOSE_MAP entries - Move orchestrate-review and validate-delivery skills from next-task to prepare-delivery in STATIC_SKILLS - plugins.txt: add both entries - Tests: update hardcoded plugin count (17 -> 19)
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Summary of ChangesHello, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request integrates two new plugins, Highlights
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here. Footnotes
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces new plugins, specifically prepare-delivery and gate-and-ship, updating the marketplace metadata, associated test counts, and documentation generation scripts to reflect these additions. The primary feedback suggests that the gate-and-ship plugin's category in marketplace.json should be changed from 'deployment' to 'productivity' for better consistency with other orchestrator plugins.
.claude-plugin/marketplace.json
Outdated
| }, | ||
| "description": "Quality gates then ship - chains /prepare-delivery then /ship in one command", | ||
| "version": "0.1.0", | ||
| "category": "deployment", |
There was a problem hiding this comment.
For consistency with other orchestrator plugins like next-task, the category for gate-and-ship should be productivity. The PR description also refers to it as an orchestrator. The deployment category is better suited for plugins that only handle deployment, like ship, whereas gate-and-ship is a workflow that includes deployment as a final step.
| "category": "deployment", | |
| "category": "productivity", |
There was a problem hiding this comment.
Fixed - changed to productivity in e0343ae.
There was a problem hiding this comment.
Pull request overview
Integrates two new workflow plugins into the AgentSys marketplace and updates supporting metadata/scripts so plugin discovery, docs generation, and CLI tests reflect the new marketplace size.
Changes:
- Added
prepare-deliveryandgate-and-shipentries to.claude-plugin/marketplace.json(17 → 19 plugins). - Updated docs-generation metadata (
CATEGORY_MAP,STATIC_SKILLS,PURPOSE_MAP) and shifted relevant skills fromnext-tasktoprepare-delivery. - Updated plugin lists and tests to reflect the new plugin count; updated site skill stats.
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
site/content.json |
Updates the displayed Skills count (38 → 40). |
scripts/plugins.txt |
Expands the plugin list to 19 entries (used for scripting/validation). |
scripts/generate-docs.js |
Adds category/purpose mappings and static skills for the new plugins; moves skills from next-task to prepare-delivery. |
__tests__/cli-subcommands.test.js |
Updates assertions for marketplace/search plugin counts (17 → 19). |
.claude-plugin/marketplace.json |
Adds the two new plugin definitions and updates the marketplace description to 19 plugins. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| { | ||
| "value": "38", | ||
| "value": "40", | ||
| "label": "Skills", | ||
| "suffix": "" |
There was a problem hiding this comment.
The Skills stat was updated to 40, but the meta.description string in this same file still says "...38 agents, 39 skills...". Please update meta.description (or have the generator update it) so the headline counts stay consistent with the stats.
There was a problem hiding this comment.
Fixed - updated meta.description from 38/39 to 47/40 in e0343ae.
| // Purpose mapping for architecture table | ||
| const PURPOSE_MAP = { | ||
| 'next-task': 'Master workflow orchestration', | ||
| 'prepare-delivery': 'Pre-ship quality gates', | ||
| 'gate-and-ship': 'Quality gates then ship', |
There was a problem hiding this comment.
prepare-delivery / gate-and-ship are added to CATEGORY_MAP/PURPOSE_MAP, but generateCommandsTable() still only emits commands listed in COMMAND_ORDER/COMMAND_SUMMARIES when command markdown files aren’t discoverable locally. Without adding these new commands there (or introducing a static commands fallback), the generated Commands table will omit /prepare-delivery and /gate-and-ship.
There was a problem hiding this comment.
Fixed - added both commands to COMMAND_ORDER and COMMAND_SUMMARIES, plus README detail sections in e0343ae.
- Change gate-and-ship category from deployment to productivity - Fix site/content.json meta.description stale counts (38/39 -> 47/40) - Add prepare-delivery and gate-and-ship to COMMAND_ORDER/COMMAND_SUMMARIES - Add command detail sections to README.md - Update commands table, skills table, and composability docs in README - Update "How Commands Work Together" with composable delivery chain
- README: fix skills count 39 -> 40 (3 places) - AGENTS.md: fix agent count 35 -> 47, skill count 32 -> 40, add /prepare-delivery and /gate-and-ship to Core Workflow commands - docs/ARCHITECTURE.md: plugin count 13 -> 19, platform count 3 -> 5, remove stale repo-map/ directory reference, fix repo-map.json label - docs/CROSS_PLATFORM.md: remove stale repo-map/ directory reference - scripts/plugins.txt: rename repo-map -> repo-intel
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 9 out of 9 changed files in this pull request and generated 3 comments.
Comments suppressed due to low confidence (3)
docs/CROSS_PLATFORM.md:294
- This shared
lib/tree omitsrepo-map/, butlib/repo-map/still exists in the repository. Please include it in the list (or clarify why it’s intentionally excluded) so the cross-platform docs match the actual library structure.
├── platform/ # Project type detection
├── drift-detect/ # Drift detection collectors
├── schemas/ # JSON schemas for validation
├── sources/ # Task source discovery
├── state/ # Workflow state management
├── types/ # TypeScript-style type definitions
AGENTS.md:173
- This section updates the skill count to 40, but the “Workflow” key skills list still omits the newly introduced workflow skills (
prepare-delivery,check-test-coverage). Please update the key-skill rows so they reflect the current workflow skill set (and align with the README skills table).
40 skills across plugins. Agents invoke skills for reusable implementation.
| Category | Key Skills |
|----------|------------|
| Workflow | `orchestrate-review`, `discover-tasks`, `validate-delivery` |
docs/ARCHITECTURE.md:27
- The Overview claims AgentSys supports five AI coding assistants, but the list only enumerates three (Claude Code, OpenCode, Codex CLI). Either add the missing two (e.g., Cursor and Kiro, as referenced elsewhere) or change the count to match the list so the documentation is internally consistent.
AgentSys supports five AI coding assistants through a unified architecture:
1. **Claude Code** - Anthropic's official CLI (primary target)
2. **OpenCode** - Multi-model AI coding assistant
3. **Codex CLI** - OpenAI's command-line interface
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| │ ├── enhance/ # Quality analyzers (agent, plugin, docs, hooks, skills) | ||
| │ ├── perf/ # Performance investigation workflow | ||
| │ ├── patterns/ # Code analysis | ||
| │ │ ├── pipeline.js # 3-phase slop detection | ||
| │ │ ├── slop-patterns.js # Pattern definitions | ||
| │ │ └── slop-analyzers.js # Multi-pass analyzers | ||
| │ ├── repo-map/ # AST repo map generation | ||
| │ ├── state/ # Workflow state | ||
| │ └── sources/ # Task source discovery |
There was a problem hiding this comment.
The architecture tree under lib/ no longer lists lib/repo-map/, but that directory still exists in the repo (and the doc still references repo-map.json later). Please re-add repo-map/ to this directory listing (or rename/remove the directory if it’s truly deprecated) to avoid misleading readers about the code layout.
| "title": "agentsys", | ||
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 38 agents, 39 skills - structured pipelines for Claude Code, OpenCode, and Codex CLI.", | ||
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 47 agents, 40 skills - structured pipelines for Claude Code, OpenCode, and Codex CLI.", | ||
| "url": "https://agent-sh.github.io/agentsys", |
There was a problem hiding this comment.
meta.description says pipelines are for “Claude Code, OpenCode, and Codex CLI”, but the stats report 5 platforms. Also, the commands list in this file doesn’t include the newly added /prepare-delivery and /gate-and-ship commands. Please update the description and commands list so the website content reflects the current supported platforms and command set.
| | Command | What it does | | ||
| |---------|--------------| | ||
| | [`/next-task`](#next-task) | Task workflow: discovery, implementation, PR, merge | | ||
| | [`/prepare-delivery`](#prepare-delivery) | Pre-ship quality gates: deslop, review, validation, docs sync | | ||
| | [`/gate-and-ship`](#gate-and-ship) | Quality gates then ship (/prepare-delivery + /ship) | |
There was a problem hiding this comment.
scripts/generate-docs.js is configured to inject readme-commands / readme-skills into README.md between <!-- GEN:START:* --> markers, but README.md currently has no such markers. As a result, gen-docs --check can’t detect drift for these tables, and future updates to COMMAND_ORDER/STATIC_SKILLS won’t be reflected automatically. Consider adding the markers around the Commands/Skills tables (or removing README.md from FILE_MAP if it’s intentionally fully manual).
- CHANGELOG: add [Unreleased] section for prepare-delivery/gate-and-ship - docs/reference/AGENTS.md: add 9 missing plugin agent sections (prepare-delivery, gate-and-ship, consult, debate, web-ctl, ship, skillers, onboard, can-i-help), fix counts (41->47 agents, 12->19 plugins) - site/content.json: add /prepare-delivery and /gate-and-ship commands, add cursor and kiro platforms
…ry in reference docs
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 11 out of 11 changed files in this pull request and generated 4 comments.
Comments suppressed due to low confidence (2)
docs/ARCHITECTURE.md:28
- This section says “supports five AI coding assistants” but the enumerated list still only contains 3 items. Please either add the missing platforms (Cursor, Kiro) to the list (and any related examples like
detectPlatform()/ state dirs) or revert the count to match what’s actually listed.
AgentSys supports five AI coding assistants through a unified architecture:
1. **Claude Code** - Anthropic's official CLI (primary target)
2. **OpenCode** - Multi-model AI coding assistant
3. **Codex CLI** - OpenAI's command-line interface
docs/CROSS_PLATFORM.md:295
- The “Shared Libraries” directory listing omits
lib/repo-map/, but that directory exists in the repo. Please include it (or rename it in the docs if the intent was to document a different path) so the documentedlib/layout is accurate.
lib/
├── config/ # Configuration management
├── cross-platform/ # Platform detection, utilities
├── enhance/ # Quality analyzer logic
├── patterns/ # 3-phase slop detection pipeline
│ ├── pipeline.js # Orchestrates phases
│ ├── slop-patterns.js # Regex patterns (HIGH certainty)
│ └── slop-analyzers.js # Multi-pass analyzers (MEDIUM)
├── platform/ # Project type detection
├── drift-detect/ # Drift detection collectors
├── schemas/ # JSON schemas for validation
├── sources/ # Task source discovery
├── state/ # Workflow state management
├── types/ # TypeScript-style type definitions
└── utils/ # Shell escape, context optimization
</details>
---
💡 <a href="/agent-sh/agentsys/new/main?filename=.github/instructions/*.instructions.md" class="Link--inTextBlock" target="_blank" rel="noopener noreferrer">Add Copilot custom instructions</a> for smarter, more guided reviews. <a href="https://docs.github.com/en/copilot/customizing-copilot/adding-repository-custom-instructions-for-github-copilot" class="Link--inTextBlock" target="_blank" rel="noopener noreferrer">Learn how to get started</a>.
site/content.json
Outdated
| "meta": { | ||
| "title": "agentsys", | ||
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 38 agents, 39 skills - structured pipelines for Claude Code, OpenCode, and Codex CLI.", | ||
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 47 agents, 40 skills - structured pipelines for Claude Code, OpenCode, and Codex CLI.", |
There was a problem hiding this comment.
meta.description still says the pipelines are for “Claude Code, OpenCode, and Codex CLI”, but this file now advertises 5 platforms (and includes Cursor/Kiro under installation.platforms). Please update the description to reflect all supported platforms (or make it platform-agnostic) so the headline doesn’t contradict the stats below.
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 47 agents, 40 skills - structured pipelines for Claude Code, OpenCode, and Codex CLI.", | |
| "description": "A modular runtime and orchestration system for AI agents. 19 plugins, 47 agents, 40 skills - structured pipelines for modern AI coding tools.", |
| │ │ └── slop-analyzers.js # Multi-pass analyzers | ||
| │ ├── repo-map/ # AST repo map generation | ||
| │ ├── state/ # Workflow state | ||
| │ └── sources/ # Task source discovery |
There was a problem hiding this comment.
The architecture tree under lib/ no longer lists lib/repo-map/, but that directory still exists in this repo (and is referenced elsewhere for repo-map.json caching/conversion). Please re-add it to the tree (with an updated description if needed) so the documented layout matches the codebase.
| │ └── sources/ # Task source discovery | |
| │ ├── sources/ # Task source discovery | |
| │ └── repo-map/ # repo-map.json caching and conversion utilities |
README.md
Outdated
| 1. **Pre-review gates** (parallel) - deslop + /simplify + test-coverage-checker | ||
| 2. **Config lint** (conditional) - agnix + /enhance when changes touch agent/skill/plugin files |
There was a problem hiding this comment.
In the /prepare-delivery description, “deslop + /simplify + test-coverage-checker” mixes agent/plugin names and implies a /simplify command, but /simplify isn’t documented elsewhere in this repo’s command list. Please clarify whether this is a built-in platform command, a separate plugin/skill, or just a descriptive verb (and adjust the formatting accordingly).
| 1. **Pre-review gates** (parallel) - deslop + /simplify + test-coverage-checker | |
| 2. **Config lint** (conditional) - agnix + /enhance when changes touch agent/skill/plugin files | |
| 1. **Pre-review gates** (parallel) - deslop + simplify + test-coverage-checker | |
| 2. **Config lint** (conditional) - agnix + enhance when changes touch agent/skill/plugin files |
| ↓ | ||
| deslop-agent cleans AI artifacts | ||
| deslop-agent + test-coverage-checker + /simplify (parallel) | ||
| ↓ |
There was a problem hiding this comment.
This workflow diagram still references “/simplify” as a parallel step. If /simplify is not an agentsys command (or not available on all platforms), this is misleading—please either link it to the correct command/plugin/skill name or add a brief note explaining what /simplify refers to.
12 files updated: - README.md, AGENTS.md: bare delivery-validator/test-coverage-checker -> prepare-delivery: prefix - docs/workflows/NEXT-TASK.md: 7 agent references prefixed - docs/CROSS_PLATFORM.md: agent table prefixed - docs/ARCHITECTURE.md, docs/ORG_ARCHITECTURE.md: plugin count 13 -> 19 - agent-docs/workflow.md, agent-docs/OPENCODE-REFERENCE.md: agent refs prefixed - docs/reference/AGENTS.md: invocation refs prefixed - site/ux-spec.md: counts 15/35/32 -> 19/47/40 (5 occurrences) - site/content.json: git-map -> repo-intel in release note - meta/skills/maintain-cross-platform/SKILL.md: orchestrate-review/validate-delivery moved from next-task to prepare-delivery attribution
11 files updated to reflect current state (19 plugins, 47 agents, 40 skills, 5 platforms): - ARCHITECTURE.md: command tables 9->20, platform sections with Cursor/Kiro, test count 1400->3445, compatibility table complete - CROSS_PLATFORM.md: command lists x3 platforms 8->20, agent count 39->47, perf model assignments fixed (4 agents opus not sonnet) - README.md (docs/): commands table 9->20, agent count 42->47, added Cursor/Kiro state dirs, repo-map.json->repo-intel.json - USAGE.md: command table 9->20 - INSTALLATION.md: verification section +8 commands - TESTING.md: /delivery-approval->/prepare-delivery, repo-map->repo-intel - ORG_ARCHITECTURE.md: +10 missing repos, date updated - CONTRIBUTING.md: 3->5 platforms, plugins/ reference updated - AGENTS.md: OpenCode state dir fixed - marketplace.json: next-task agent count 14->8, version 1.0.0->1.1.0 - site/content.json: platform description, test counts
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 22 out of 22 changed files in this pull request and generated 3 comments.
Comments suppressed due to low confidence (1)
AGENTS.md:174
- The “Workflow” key-skill list hasn’t been updated to reflect the new skills introduced/moved in this PR (e.g.,
prepare-deliveryandcheck-test-coverage). As written, it still only listsorchestrate-review,discover-tasks,validate-delivery. Please update this table so it matches the README’s updated workflow skill set.
40 skills across plugins. Agents invoke skills for reusable implementation.
| Category | Key Skills |
|----------|------------|
| Workflow | `orchestrate-review`, `discover-tasks`, `validate-delivery` |
| Enhancement | `enhance-*` (9 skills for plugins, agents, docs, prompts, hooks) |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
|
|
||
| <p align="center"> | ||
| <b>19 plugins · 47 agents · 39 skills (across all repos) · 30k lines of lib code · 3,583 tests · 5 platforms</b><br> | ||
| <b>19 plugins · 47 agents · 40 skills (across all repos) · 30k lines of lib code · 3,583 tests · 5 platforms</b><br> |
There was a problem hiding this comment.
The README still advertises "3,583 tests", but other updated sources in this PR (e.g., site/content.json stats + docs/ARCHITECTURE.md) were changed to 3,445. Please update the test count here (and the other README occurrence under “Testing”) so the headline metrics stay consistent.
| <b>19 plugins · 47 agents · 40 skills (across all repos) · 30k lines of lib code · 3,583 tests · 5 platforms</b><br> | |
| <b>19 plugins · 47 agents · 40 skills (across all repos) · 30k lines of lib code · 3,445 tests · 5 platforms</b><br> |
| | `flow.json` | `{state-dir}/` (worktree) | Which phase you're in | | ||
| | `preference.json` | `{state-dir}/sources/` | Cached task source preference | | ||
| | `repo-map.json` | `{state-dir}/` | Cached AST repo map | | ||
| | `repo-intel.json` | `{state-dir}/` | Cached symbol map | |
There was a problem hiding this comment.
This state file row was changed to repo-intel.json as the “Cached symbol map”, but the code persists both {state-dir}/repo-intel.json (raw agent-analyzer output) and {state-dir}/repo-map.json (converted symbol map used by the repo-map cache). Please update this row to reflect the actual file(s) and purposes to avoid misleading users.
| | `repo-intel.json` | `{state-dir}/` | Cached symbol map | | |
| | `repo-intel.json` | `{state-dir}/` | Raw agent-analyzer output used to build symbol map | | |
| | `repo-map.json` | `{state-dir}/` | Converted symbol map used by the repo-map cache | |
| ├── planning-agent.md | ||
| ├── implementation-agent.md | ||
| ├── delivery-validator.md | ||
| ├── prepare-delivery:delivery-validator.md |
There was a problem hiding this comment.
This example uses an agent filename containing a colon (prepare-delivery:delivery-validator.md). On Windows, : is not a valid filename character, and the installer copies agent files verbatim into ~/.config/opencode/agents/ (no renaming). Please change the example to use a filesystem-safe filename (e.g., delivery-validator.md) and keep the plugin prefix only in the agent’s frontmatter/name if needed.
| ├── prepare-delivery:delivery-validator.md | |
| ├── delivery-validator.md |
Summary
New plugins
Composability
Test plan