Convert code-review workflow to native skill packaging#1941
Merged
Conversation
🤖 Augment PR SummarySummary: Migrates the BMM code-review workflow into a native skill ( Changes: Renames the workflow directory, updates workflow frontmatter/path assumptions for the new layout, and repoints agent menu triggers to the new location. 🤖 Was this summary useful? React with 👍 or 👎 |
![augmentcode[bot]](https://avatars.githubusercontent.com/in/1027498?s=60&v=4){kind=small}
alexeyv
force-pushed
the
convert-code-review-skill
branch
from
b114de0 to
d206560
Compare
📝 WalkthroughWalkthroughThe PR converts the bmad-code-review workflow into a native skill by creating a skill metadata structure (SKILL.md and manifest), removing the old workflow manifest, updating path references, and modifying two agent files to reference the skill via "skill:bmad-code-review" instead of filesystem paths. Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
📝 Coding Plan
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (2)
src/bmm/workflows/4-implementation/bmad-code-review/SKILL.md (1)
3-3: Description text is over-coupled to literal trigger phrasing.Line 3 hard-codes specific English utterances (
"run code review","review this code"). That will drift from actual trigger sources (agent menus/module-help) and is brittle for non-English sessions. Keep SKILL metadata capability-focused and keep trigger logic in dispatch configs.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/bmm/workflows/4-implementation/bmad-code-review/SKILL.md` at line 3, The SKILL metadata's description currently hard-codes example trigger phrases ("run code review", "review this code"); remove those literal utterances from the description field in SKILL.md and replace with a capability-focused sentence describing what the skill does (e.g., "Perform adversarial code review to identify issues and suggest fixes"), leaving trigger phrasing to dispatch/trigger configs; update the description key in SKILL.md (the 'description' value) only—do not add trigger logic here and ensure no language-specific examples remain.src/bmm/agents/dev.agent.yaml (1)
37-37: Hardcoded skill ID is duplicated across agent menus.
skill:bmad-code-reviewis now repeated in multiple agents. Centralize this mapping (single source of truth) to avoid future drift when IDs or routing change.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/bmm/agents/dev.agent.yaml` at line 37, The exec entry "skill:bmad-code-review" is hardcoded and duplicated across agent menus; centralize this mapping by moving the skill ID into a single source (e.g., a shared skills mapping file or YAML anchor/alias) and reference that centralized symbol from each agent instead of embedding "skill:bmad-code-review" directly; update the dev.agent.yaml exec field to use the shared identifier (the centralized key you create) and change other agent YAMLs to reference the same key so all agents use one canonical mapping (search for exec: "skill:bmad-code-review" to locate all occurrences and replace them with the centralized reference).
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/bmm/workflows/4-implementation/bmad-code-review/SKILL.md`:
- Line 6: The skill entrypoint currently launches the workflow directly ("Follow
the instructions in workflow.md.") and must not auto-run high-impact fixes
without confirmation; modify the entry logic that invokes the workflow to
present an explicit modal/CLI prompt or flag check distinguishing read-only
(scan-only) vs fix-mode (auto-fix) before proceeding, require the user to
affirmatively choose fix-mode (e.g., `--apply` or an interactive "Apply fixes?
(yes/no)") and clearly log the chosen mode; update any references in SKILL.md
and the workflow invocation code to document and enforce this confirmation step.
---
Nitpick comments:
In `@src/bmm/agents/dev.agent.yaml`:
- Line 37: The exec entry "skill:bmad-code-review" is hardcoded and duplicated
across agent menus; centralize this mapping by moving the skill ID into a single
source (e.g., a shared skills mapping file or YAML anchor/alias) and reference
that centralized symbol from each agent instead of embedding
"skill:bmad-code-review" directly; update the dev.agent.yaml exec field to use
the shared identifier (the centralized key you create) and change other agent
YAMLs to reference the same key so all agents use one canonical mapping (search
for exec: "skill:bmad-code-review" to locate all occurrences and replace them
with the centralized reference).
In `@src/bmm/workflows/4-implementation/bmad-code-review/SKILL.md`:
- Line 3: The SKILL metadata's description currently hard-codes example trigger
phrases ("run code review", "review this code"); remove those literal utterances
from the description field in SKILL.md and replace with a capability-focused
sentence describing what the skill does (e.g., "Perform adversarial code review
to identify issues and suggest fixes"), leaving trigger phrasing to
dispatch/trigger configs; update the description key in SKILL.md (the
'description' value) only—do not add trigger logic here and ensure no
language-specific examples remain.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 5767030b-c6f9-4273-afe1-f5b845ff774b
⛔ Files ignored due to path filters (1)
src/bmm/module-help.csvis excluded by!**/*.csv
📒 Files selected for processing (8)
src/bmm/agents/dev.agent.yamlsrc/bmm/agents/quick-flow-solo-dev.agent.yamlsrc/bmm/workflows/4-implementation/bmad-code-review/SKILL.mdsrc/bmm/workflows/4-implementation/bmad-code-review/bmad-skill-manifest.yamlsrc/bmm/workflows/4-implementation/bmad-code-review/checklist.mdsrc/bmm/workflows/4-implementation/bmad-code-review/discover-inputs.mdsrc/bmm/workflows/4-implementation/bmad-code-review/workflow.mdsrc/bmm/workflows/4-implementation/code-review/bmad-skill-manifest.yaml
💤 Files with no reviewable changes (1)
- src/bmm/workflows/4-implementation/code-review/bmad-skill-manifest.yaml
| description: 'Perform adversarial code review finding specific issues. Use when the user says "run code review" or "review this code"' | ||
| --- | ||
|
|
||
| Follow the instructions in [workflow.md](workflow.md). |
There was a problem hiding this comment.
Skill entrypoint does not warn about write side effects.
Line 6 directly launches a workflow that can auto-fix code and update story/sprint files. That is high-impact behavior for a command users may invoke as “just review.” Add an explicit read-only vs fix-mode confirmation at entry.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/bmm/workflows/4-implementation/bmad-code-review/SKILL.md` at line 6, The
skill entrypoint currently launches the workflow directly ("Follow the
instructions in workflow.md.") and must not auto-run high-impact fixes without
confirmation; modify the entry logic that invokes the workflow to present an
explicit modal/CLI prompt or flag check distinguishing read-only (scan-only) vs
fix-mode (auto-fix) before proceeding, require the user to affirmatively choose
fix-mode (e.g., `--apply` or an interactive "Apply fixes? (yes/no)") and clearly
log the chosen mode; update any references in SKILL.md and the workflow
invocation code to document and enforce this confirmation step.
This was referenced Mar 14, 2026
This was referenced Apr 1, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
src/bmm/workflows/4-implementation/code-reviewinto native skill directorysrc/bmm/workflows/4-implementation/bmad-code-reviewtype: skilland addSKILL.mdmodule-help.csv,dev.agent.yaml, andquick-flow-solo-dev.agent.yamlValidation
node tools/cli/bmad-cli.js install --directory /Users/alex/src/bmad --modules bmm --tools claude-code --yes/Users/alex/src/bmad/.claude/skills/bmad-code-review/_bmad/_config/skill-manifest.csv; absent from_bmad/_config/workflow-manifest.csvafter clean reinstall passdiscover-inputs.mdhash parity PASS;checklist.mdhash parity PASSnpm test