@delete101020/aidlc 4.0.0
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/AUTOMATION.md +148 -0
- package/README.md +684 -0
- package/dist/assets/annotate-artifact.skill.md +121 -0
- package/dist/assets/epic-context.skill.md +55 -0
- package/dist/bundle.js +49470 -0
- package/dist/templates/ainative/agents/native-engineer.md +66 -0
- package/dist/templates/ainative/agents/native-operator.md +71 -0
- package/dist/templates/ainative/agents/native-originator.md +56 -0
- package/dist/templates/ainative/agents/native-product-owner.md +68 -0
- package/dist/templates/ainative/agents/native-reviewer.md +65 -0
- package/dist/templates/ainative/agents/native-verifier.md +58 -0
- package/dist/templates/ainative/artifacts/build-plan.md +56 -0
- package/dist/templates/ainative/artifacts/implement.md +55 -0
- package/dist/templates/ainative/artifacts/intent.md +50 -0
- package/dist/templates/ainative/artifacts/maintain.md +85 -0
- package/dist/templates/ainative/artifacts/review.md +68 -0
- package/dist/templates/ainative/artifacts/spec.md +69 -0
- package/dist/templates/ainative/artifacts/verify.md +55 -0
- package/dist/templates/ainative/skills/native-build-plan.md +74 -0
- package/dist/templates/ainative/skills/native-implement.md +62 -0
- package/dist/templates/ainative/skills/native-intent.md +50 -0
- package/dist/templates/ainative/skills/native-maintain.md +94 -0
- package/dist/templates/ainative/skills/native-review.md +82 -0
- package/dist/templates/ainative/skills/native-spec.md +83 -0
- package/dist/templates/ainative/skills/native-verify.md +60 -0
- package/dist/templates/sdlc/agents/designer.md +87 -0
- package/dist/templates/sdlc/agents/developer.md +178 -0
- package/dist/templates/sdlc/agents/example-with-skills.md +59 -0
- package/dist/templates/sdlc/agents/orchestrator.md +186 -0
- package/dist/templates/sdlc/agents/po.md +115 -0
- package/dist/templates/sdlc/agents/qa.md +141 -0
- package/dist/templates/sdlc/agents/tech-lead.md +124 -0
- package/dist/templates/sdlc/artifacts/design.md +130 -0
- package/dist/templates/sdlc/artifacts/execute-test.md +64 -0
- package/dist/templates/sdlc/artifacts/implement.backend.md +96 -0
- package/dist/templates/sdlc/artifacts/implement.md +70 -0
- package/dist/templates/sdlc/artifacts/implement.web-react.md +109 -0
- package/dist/templates/sdlc/artifacts/implement.web.md +99 -0
- package/dist/templates/sdlc/artifacts/plan.md +97 -0
- package/dist/templates/sdlc/artifacts/prototype.md +141 -0
- package/dist/templates/sdlc/artifacts/test-cases.md +47 -0
- package/dist/templates/sdlc/artifacts/test-plan.md +100 -0
- package/dist/templates/sdlc/artifacts/unit-test.md +32 -0
- package/dist/templates/sdlc/profiles/agile-lite.yaml +29 -0
- package/dist/templates/sdlc/profiles/hybrid.yaml +43 -0
- package/dist/templates/sdlc/profiles/iso-ieee.yaml +67 -0
- package/dist/templates/sdlc/profiles/none.yaml +21 -0
- package/dist/templates/sdlc/skills/advance-epic.md +77 -0
- package/dist/templates/sdlc/skills/analyze-requirements.md +279 -0
- package/dist/templates/sdlc/skills/discovery-gate.md +118 -0
- package/dist/templates/sdlc/skills/execute-test.md +74 -0
- package/dist/templates/sdlc/skills/generate-test-cases.md +77 -0
- package/dist/templates/sdlc/skills/implement.md +74 -0
- package/dist/templates/sdlc/skills/prd.md +114 -0
- package/dist/templates/sdlc/skills/prototype.md +128 -0
- package/dist/templates/sdlc/skills/tech-design.md +134 -0
- package/dist/templates/sdlc/skills/test-plan.md +115 -0
- package/dist/templates/sdlc/skills/test-report.md +38 -0
- package/dist/templates/sdlc/skills/unit-test.md +43 -0
- package/dist/templates/sdlc/validators/ci.mjs +49 -0
- package/dist/templates/sdlc/validators/traceability.mjs +244 -0
- package/dist/templates/speckit/agents/analyst.md +42 -0
- package/dist/templates/speckit/agents/speckit-developer.md +36 -0
- package/dist/templates/speckit/agents/speckit-qa.md +38 -0
- package/dist/templates/speckit/agents/speckit-tech-lead.md +40 -0
- package/dist/templates/speckit/artifacts/analyze.md +48 -0
- package/dist/templates/speckit/artifacts/clarify.md +64 -0
- package/dist/templates/speckit/artifacts/plan.md +57 -0
- package/dist/templates/speckit/artifacts/specify.md +64 -0
- package/dist/templates/speckit/artifacts/tasks.md +33 -0
- package/dist/templates/speckit/skills/analyze.md +50 -0
- package/dist/templates/speckit/skills/clarify.md +31 -0
- package/dist/templates/speckit/skills/plan.md +56 -0
- package/dist/templates/speckit/skills/specify.md +52 -0
- package/dist/templates/speckit/skills/speckit-implement.md +33 -0
- package/dist/templates/speckit/skills/tasks.md +38 -0
- package/dist/tools/epic-memory-hook.mjs +78 -0
- package/dist/tools/epic-memory.mjs +130 -0
- package/dist/tools/md-to-html.mjs +478 -0
- package/dist/tools/vendor/marked.esm.mjs +2432 -0
- package/dist/vendor/annotron/LICENSE +21 -0
- package/dist/vendor/annotron/README.md +213 -0
- package/dist/vendor/annotron/bin/annotron +467 -0
- package/dist/vendor/annotron/commands/annotron.md +11 -0
- package/dist/vendor/annotron/hooks/annotron-hook.sh +38 -0
- package/dist/vendor/annotron/hooks/hooks.json +50 -0
- package/dist/vendor/annotron/package.json +72 -0
- package/dist/vendor/annotron/skills/annotron-review/SKILL.md +86 -0
- package/dist/vendor/annotron/src/chrome.html +2029 -0
- package/dist/vendor/annotron/src/mdRender.js +208 -0
- package/dist/vendor/annotron/src/sdk.js +1042 -0
- package/dist/vendor/annotron/src/server.js +780 -0
- package/package.json +60 -0
|
@@ -0,0 +1,86 @@
|
|
|
1
|
+
---
|
|
2
|
+
description: Open agent-generated HTML artifacts in a local browser editor so the user can
|
|
3
|
+
annotate elements/text and send feedback; receive that feedback, apply it, and iterate
|
|
4
|
+
until the user finalizes. Use whenever the user wants to review or refine an HTML artifact.
|
|
5
|
+
---
|
|
6
|
+
|
|
7
|
+
## First-time setup
|
|
8
|
+
|
|
9
|
+
Before running annotron for the first time in a project, check if `.claude/settings.json`
|
|
10
|
+
already allows `Bash(curl:*)` and `Bash(annotron:*)`. If not, ask the user:
|
|
11
|
+
|
|
12
|
+
> "annotron cần tự động chạy `curl` và `annotron poll` để nhận feedback mà không cần bạn
|
|
13
|
+
> approve từng lần. Cho phép thêm auto-allow vào `.claude/settings.json` không?"
|
|
14
|
+
|
|
15
|
+
If yes, create/update `.claude/settings.json`:
|
|
16
|
+
```json
|
|
17
|
+
{
|
|
18
|
+
"permissions": {
|
|
19
|
+
"allow": ["Bash(curl:*)", "Bash(annotron:*)", "Read(/private/tmp/*)"]
|
|
20
|
+
}
|
|
21
|
+
}
|
|
22
|
+
```
|
|
23
|
+
|
|
24
|
+
---
|
|
25
|
+
|
|
26
|
+
When the user wants to review or refine an HTML artifact you produced:
|
|
27
|
+
|
|
28
|
+
1. **Open it**: run `annotron <path-to-artifact.html>`.
|
|
29
|
+
This starts the background server (if not running), registers the file, and opens the
|
|
30
|
+
review editor in the user's browser. Print the editor URL for the user.
|
|
31
|
+
|
|
32
|
+
2. **Wait for feedback**: run `annotron poll <path>`.
|
|
33
|
+
This blocks until the user sends feedback. Output is JSON with:
|
|
34
|
+
- `items[]` — each has `kind` (element | text), `selector`, `text`, `note`, and an
|
|
35
|
+
optional `images[]` (images attached to that specific annotation's note/reply box)
|
|
36
|
+
- `message` — freeform message from the user
|
|
37
|
+
- `images[]` — top-level images attached to the composer message (or a single annotation reply)
|
|
38
|
+
- Every image (top-level or per-item) has `name` and `path` (an absolute file path).
|
|
39
|
+
`Read` each `path` to see the image before applying the feedback.
|
|
40
|
+
- `finalized: true` — signals the user is done; skip to step 5
|
|
41
|
+
|
|
42
|
+
3. **Apply the feedback**: edit the HTML file using the selector/text + note from each item.
|
|
43
|
+
Each item in `items[]` now includes a server-assigned `id` field (e.g. `ann_xxx`) you can
|
|
44
|
+
use to reply to a specific annotation thread.
|
|
45
|
+
|
|
46
|
+
**Your steps are mirrored automatically.** A bundled `PostToolUse`/`PreToolUse` hook streams
|
|
47
|
+
every tool call (Read/Edit/Bash/…) into the browser sidebar as CLI-like steps, and reflects
|
|
48
|
+
idle/permission status — you do not need to narrate. You may still post a high-level milestone
|
|
49
|
+
with `annotron progress <path> "…"` if it helps, but it is optional now.
|
|
50
|
+
|
|
51
|
+
**Remote permission approval (optional, user-controlled).** If the user turns on *Remote approve*
|
|
52
|
+
in the browser, the hook routes each tool's permission prompt to the browser; the user clicks
|
|
53
|
+
Allow / Allow-always / Deny and the decision comes back to the CLI. If a tool is denied you'll get
|
|
54
|
+
a deny reason — respect it, don't retry the same action. If the browser doesn't answer in time the
|
|
55
|
+
prompt falls back to the terminal. This is transparent to you; no special commands needed.
|
|
56
|
+
|
|
57
|
+
**Cancellation is enforced automatically.** annotron ships a `PreToolUse` hook: if the user
|
|
58
|
+
clicks **Cancel** while you are mid-round, your next tool call (Edit/Write/Read/Bash, except
|
|
59
|
+
`annotron` commands) is denied with a message telling you to stop. When you see that denial,
|
|
60
|
+
do not retry other edits — reply with what you did and did not finish
|
|
61
|
+
(`annotron poll <path> --reply "Stopped — cancelled. Applied X, skipped Y."`) and wait for the
|
|
62
|
+
next feedback. The cancel flag resets automatically when the user sends new feedback.
|
|
63
|
+
(You can also proactively `annotron check <path>` between steps; it prints `{"cancelled":true}`
|
|
64
|
+
— but the hook makes this optional.)
|
|
65
|
+
|
|
66
|
+
4. **Tell the user what changed**: run `annotron poll <path> --reply "..."`.
|
|
67
|
+
This posts an agent message to the general conversation log, then re-arms the poll.
|
|
68
|
+
To reply directly to a specific annotation thread, pass `--annotation-id`:
|
|
69
|
+
`annotron poll <path> --reply "message" --annotation-id ann_xxx`
|
|
70
|
+
The browser shows that reply inline in the annotation's thread. A reply without
|
|
71
|
+
`--annotation-id` posts to the general conversation log.
|
|
72
|
+
Saving the file triggers an automatic live-reload in the browser.
|
|
73
|
+
|
|
74
|
+
5. **Repeat steps 2–4** until the poll returns `"finalized": true`.
|
|
75
|
+
At that point the user has written the confirmed result into the file — the loop is done.
|
|
76
|
+
|
|
77
|
+
## Tips
|
|
78
|
+
|
|
79
|
+
- The `bin/` directory containing `annotron` is on your PATH while this plugin is active.
|
|
80
|
+
- Keep replies short and action-focused ("Updated h1 colour, fixed revenue figure, added footer.").
|
|
81
|
+
- Never edit the file while a poll is in flight — wait for the feedback JSON first.
|
|
82
|
+
- Annotations are persisted automatically in `<artifact>.annotron.json` beside the HTML file.
|
|
83
|
+
- Pasted/uploaded images are saved to `.annotron-uploads/` beside the artifact; feedback references them by absolute `path` — `Read` those paths to view the images.
|
|
84
|
+
- Post progress liberally with `annotron progress <file> "…"` — one short line per step. Keep it action-focused ("Reading X", "Updating Y").
|
|
85
|
+
- `annotron check <file>` is cheap; run it between edits so a mid-flight cancel is honored quickly.
|
|
86
|
+
- To reply to a specific annotation thread: `annotron poll <file> --reply "msg" --annotation-id <id>` — the browser shows the reply inline in that annotation's thread.
|