alexsoyes · Nov 2, 2026 · Nov 2, 2026 · Nov 2, 2026 · Nov 3, 2026 · Nov 3, 2026
@@ -28,7 +28,7 @@
     {
       "name": "aidd-vcs",
       "source": "./plugins/aidd-vcs",
-      "description": "External artifacts: commit, pull-request, release-tag, issue-create",
+      "description": "External artifacts: commit, pull-request, release-tag, issue-create, answer-comments",
       "strict": true,
       "metadata": {
         "recommended": true

@@ -11,7 +11,7 @@ Unify **engineering teams** around **standardized workflows** and **shared best
 🧱 **IDE agnostic** · 🏗️ **Legacy systems** · 🌱 **Token-optimized** · 🇫🇷 **Made in France**
 
 <p>
-  <!--counts:start--><kbd>9 plugins</kbd> · <kbd>52 skills</kbd> · <kbd>4 agents</kbd><!--counts:end-->
+  <!--counts:start--><kbd>9 plugins</kbd> · <kbd>53 skills</kbd> · <kbd>4 agents</kbd><!--counts:end-->
 </p>
 
 [![Open Source](https://img.shields.io/badge/Open_Source-Yes-yellow?logo=open-source-initiative&logoColor=white)](https://opensource.org/)
@@ -149,7 +149,7 @@ Plan, build, review, test, debug
 <a href="plugins/aidd-vcs/README.md">
 <img align="right" alt="stable" src="https://img.shields.io/badge/stable-2da44e?style=flat-square" />
 <h3>🌿 aidd-vcs</h3>
-6 skills<br />
+7 skills<br />
 Commits, pull requests, release tags, issues
 </a>
 </td>

@@ -103,6 +103,7 @@ Version-control workflows: repo init, commit, pull/merge request, release tag, i
 | `03-release-tag`  | Cut a semver release with annotated tag and notes                              | `01-release-tag`                        |
 | `04-issue-create` | Create an issue in the configured ticketing tool                               | `01-issue-create`                       |
 | `05-resolve-conflict` | Resolve deterministic conflicts or approved choices | `01-resolve` |
+| `06-answer-comments` | Answer review comments, resolve settled threads, and propose learnings | `01-answer-comments`, `02-propose-learnings` |
 
 ## 🎼 aidd-orchestrator
 

@@ -2,7 +2,7 @@
   "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
   "name": "aidd-vcs",
   "version": "2.5.0",
-  "description": "External artifacts: repo-init, commit, pull-request, release-tag, issue-create, resolve-conflict",
+  "description": "External artifacts: repo-init, commit, pull-request, release-tag, issue-create, resolve-conflict, answer-comments",
   "author": {
     "name": "AI-Driven Dev",
     "url": "https://github.com/ai-driven-dev"
@@ -13,7 +13,8 @@
     "./skills/02-pull-request",
     "./skills/03-release-tag",
     "./skills/04-issue-create",
-    "./skills/05-resolve-conflict"
+    "./skills/05-resolve-conflict",
+    "./skills/06-answer-comments"
   ],
   "keywords": [
     "git",

@@ -14,6 +14,7 @@ Auto-generated index of skills, agents, references and assets shipped by the `ai
   - [`skills/03-release-tag`](#skills03-release-tag)
   - [`skills/04-issue-create`](#skills04-issue-create)
   - [`skills/05-resolve-conflict`](#skills05-resolve-conflict)
+  - [`skills/06-answer-comments`](#skills06-answer-comments)
 
 ---
 
@@ -80,3 +81,14 @@ Auto-generated index of skills, agents, references and assets shipped by the `ai
 | `assets` | [resolution-table.md](skills/05-resolve-conflict/assets/resolution-table.md) | - |
 | `-` | [SKILL.md](skills/05-resolve-conflict/SKILL.md) | `Resolves deterministic Git conflicts or approved choices. Use when the user wants to resolve a merge, rebase, or cherry-pick conflict. Not for deciding between competing implementations or committing changes.` |
 
+#### `skills/06-answer-comments`
+
+| Group | File | Description |
+|-------|------|---|
+| `actions` | [01-answer-comments.md](skills/06-answer-comments/actions/01-answer-comments.md) | - |
+| `actions` | [02-propose-learnings.md](skills/06-answer-comments/actions/02-propose-learnings.md) | - |
+| `assets` | [comments-summary-template.md](skills/06-answer-comments/assets/comments-summary-template.md) | - |
+| `references` | [comment-kinds.md](skills/06-answer-comments/references/comment-kinds.md) | - |
+| `references` | [learnings.md](skills/06-answer-comments/references/learnings.md) | - |
+| `-` | [SKILL.md](skills/06-answer-comments/SKILL.md) | `Answers pull or merge request review comments, resolves settled threads, and proposes learnings. Use when the user wants review feedback answered or pending decisions summarized. Not for reviewing a diff or fixing code.` |
+
@@ -0,0 +1,27 @@
+---
+name: 06-answer-comments
+description: Answers pull or merge request review comments, resolves settled threads, and proposes learnings. Use when the user wants review feedback answered or pending decisions summarized. Not for reviewing a diff or fixing code.
+argument-hint: pull request | merge request
+---
+
+# Answer Comments
+
+```mermaid
+flowchart LR
+  request([pull or merge request]) --> answer-comments --> propose-learnings --> summary([summary updated])
+  summary -.->|new reply or pushed commit| request
+```
+
+## Actions
+
+Run the flow above. Read only the next action file.
+
+| Action | Does |
+| --- | --- |
+| answer-comments | answer open review threads and resolve the settled ones |
+| propose-learnings | propose lasting feedback and update the summary |
+
+## Transversal rules
+
+- Reach the pull or merge request through the VCS tool `aidd_docs/memory/vcs.md` names, else the one the remote points to.
+- Read code and history as evidence; never modify source, commit, push, or merge the request.
@@ -0,0 +1,23 @@
+# 01 - Answer Comments
+
+Reply to each open review comment on a pull or merge request, and resolve the settled ones.
+
+## Input
+
+The pull or merge request, and the commits pushed since its comments.
+
+## Output
+
+Each open thread replied to, resolved or left open per its kind, and the `user-call` and `todo` threads named.
+
+## Process
+
+1. **Collect.** Read the open review threads, skipping those still waiting per [comment-kinds.md](../references/comment-kinds.md).
+2. **Sort.** Give each comment one kind per `comment-kinds.md`.
+3. **Reply.** Reply in each thread as its kind says.
+4. **Resolve.** Resolve or leave open each thread as its kind says.
+5. **Hand over.** Name all remaining `user-call` and `todo` threads with their links.
+
+## Rules
+
+- Include skipped waiting threads in the handover, even when no reply was posted.
@@ -0,0 +1,22 @@
+# 02 - Propose Learnings
+
+Propose the feedback that should outlive the pull or merge request, and post the summary.
+
+## Input
+
+The threads answer-comments replied to, and the ones it handed over.
+
+## Output
+
+One summary comment on the pull or merge request, shaped by [comments-summary-template.md](../assets/comments-summary-template.md) and kept up to date across runs.
+
+## Process
+
+1. **Keep.** Sort the feedback per [learnings.md](../references/learnings.md).
+2. **Fill.** Fill `comments-summary-template.md`.
+3. **Post.** Post it on the first run; on a later one, edit the comment opening with `## Review comments` instead.
+4. **Offer.** If there are learnings, propose running the skill that captures durable project learnings with this summary as input.
+
+## Rules
+
+- Keep learnings as proposals; write no rule, memory or skill.
@@ -0,0 +1,17 @@
+## Review comments
+
+Done: {{n}}{{ (short shas), when there are any}}. Answered: {{n}}.
+
+### FAQ: your call
+
+1. **{{question}}** ([thread]({{url}})): {{option a}} / {{option b}}
+
+### To do
+
+- {{change}} ([thread]({{url}}))
+
+### Learnings to keep
+
+- {{lesson}} ([thread]({{url}}))
+
+<!-- Fill every {{...}}, one line per item. Drop a section left empty, and this comment. Keep it short enough to read on a phone. -->
@@ -0,0 +1,22 @@
+# Comment kinds
+
+The kind each review comment gets, and what its thread receives.
+
+## Kinds
+
+| Kind | When | Reply | Thread |
+| --- | --- | --- | --- |
+| `done` | a change it asked for is in a commit pushed after it | what changed, and the short commit sha | resolved |
+| `answer` | a question the code, docs, tests or history answer | the answer, and the file or commit it rests on | resolved |
+| `user-call` | a choice only the user can settle: intent, trade-off, scope, taste | the question in one sentence and its options, ending `Answer here.` | open |
+| `user-call`, answered | the user has stated their decision, even if implementation remains | `Decision:` and the user's choice | resolved |
+| `todo` | a change asked for, not pushed yet | the change, ending `To do.` | open |
+
+## Rules
+
+| Rule | Why |
+| --- | --- |
+| Between `answer` and `user-call`, pick `user-call` | a guessed answer reads as decided |
+| Reply in one or two lines, in the comment's language, never quoting it back | the thread stays readable |
+| Keep `Answer here.` and `To do.` verbatim, in English | a later run reads them to know which threads still wait |
+| Skip a thread ending `Answer here.`, or `To do.` with no commit pushed since | it still waits, whatever account posted it |
@@ -0,0 +1,11 @@
+# Learnings
+
+Which review feedback outlives its pull or merge request.
+
+| Feedback | Becomes |
+| --- | --- |
+| a preference stated as a general one, such as "docs too long" | a learning |
+| the same remark on several files | one learning, linking every thread |
+| a decision another change would face again | a learning |
+| what a project rule or memory already says | that rule named as not followed, no new learning |
+| a one-off change | nothing |
@@ -229,7 +229,7 @@ test("sweeping the repository's own plugins/ tree yields zero violations", () =>
   assert.deepEqual(violations, []);
   // A sweep that never actually exercised a skill with action files would pass the same way —
   // this pins the sweep to the measured count so it cannot go vacuously green.
-  assert.equal(skillsWithActions, 49);
+  assert.equal(skillsWithActions, 50);
 });
 
 test("a table that declares no action column cites nothing, whatever its cells read", () => {