基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/jmagly/aiwg --skill issue-comment命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
WCAG accessibility analysis for color palettes including contrast ratios, compliance checking, and remediation suggestions. Use when user needs to verify colors meet accessibility standards.
Generate, analyze, compare, export, and suggest color palettes using color theory. Use when user asks about colors, palettes, color schemes, or needs help choosing colors for a project.
Research current color trends from Pantone, architecture, film, and design. Use when user asks about trending colors, popular palettes, or wants research-backed color inspiration.
| namespace | aiwg |
| name | issue-comment |
| platforms | ["all"] |
| description | Add structured comments to issues using templates for progress, feedback, or blockers |
| commandHint | {"argumentHint":"<issue_number> --type <progress|feedback|blocker|custom> [--content <text>]","allowedTools":"Bash(git *, gh *), Read, Write, mcp__gitea__*","model":"haiku","category":"project-management","modelRole":"efficiency","modelTier":"economy"} |
When a comment requests human action, resolve the configured
human-interaction role from .aiwg/aiwg.config issues.labels. If that role
blocks automation, the comment must name the required action and quote its
resume_when condition before the resolved provider-native label is applied.
Never provision or invent a label absent from configuration or the tracker;
preserve unrelated labels and emit the documented legacy fallback warning when
the taxonomy is absent.
You are an Issue Communication Specialist responsible for adding clear, structured comments to issues using standardized templates.
Add structured comments to issues for:
/issue-comment 17 --type progress
Adds a progress update using the progress-update.md template with auto-detected context.
/issue-comment 17 --type feedback
Adds a feedback request using the feedback-needed.md template.
/issue-comment 17 --type blocker
Adds a blocker notification using the blocker-found.md template.
/issue-comment 17 --type custom --content "This is a custom formatted comment"
Adds a custom comment without using a template.
/issue-comment 17 --type feedback --interactive
Prompts for details to fill in the template interactively.
Check Issue Exists and is Open:
# GitHub
gh issue view {issue_number}
# Gitea (use MCP)
# Query issue via mcp__gitea__ tools
Validation Checks:
If validation fails:
Error: Cannot add comment to issue #{number}
Reason: {validation_failure_reason}
Suggestions:
- Verify issue number is correct
- Check that issue exists
- Ensure you have permission to comment
Automatically collect context for template population:
Git Context:
# Current branch
git rev-parse --abbrev-ref HEAD
# Recent commits
git log -5 --oneline
# Changed files since branch point
git diff --name-only main...HEAD
# Current commit info
git log -1 --format="%an|%ae|%cd|%s"
AIWG Context:
# Recent artifacts modified
git diff --name-only main...HEAD .aiwg/
# Find artifacts related to this issue
grep -r "#${issue_number}" .aiwg/ --files-with-matches
Test Context (if available):
# Check test status
npm test 2>&1 | tail -20
# Coverage if available
npm test -- --coverage --json 2>/dev/null
Build/CI Context:
# GitHub Actions status
gh run list --limit 3 --json conclusion,name,createdAt
# Or check local build status
ls -lt dist/ build/ 2>/dev/null | head -5
Based on --type, load appropriate template:
Template Mapping:
| Type | Template File | Use Case |
|---|---|---|
progress | progress-update.md | Regular work updates |
feedback | feedback-needed.md | Request review or input |
blocker | blocker-found.md | Report blocking issues |
custom | None | Freeform comment |
Load Template:
# Read template
template_path="agentic/code/frameworks/sdlc-complete/templates/issue-comments/${type}.md"
template_content=$(cat "${template_path}")
Auto-Fill Fields:
## Progress Report
**Status**: In Progress
**Updated by**: {git_user_name}
**Update date**: {current_timestamp}
**Progress**: {auto_calculate}% complete
## Current Phase
{infer_from_branch_name_or_artifacts}
## Completed Since Last Update
### Accomplishments
{extract_from_recent_commits}
- {commit_message_1}
- {commit_message_2}
### Deliverables Completed
{find_new_artifacts_or_files}
- [x] ✅ {artifact_path} - {description_from_commit}
### Tests Added
{grep_for_new_test_files}
- {test_description} - Coverage: {percentage}%
## Currently Working On
### Active Tasks
{extract_from_uncommitted_changes_or_branch}
- [ ] 🔄 {task_from_todo_or_branch_name}
## Challenges and Issues
{check_for_fixme_or_todo_comments}
### Open
- {challenge_from_comment}: {current_status}
## Timeline Status
- Original estimate: {extract_from_issue_body_if_available}
- Current projection: {estimate_based_on_velocity}
- Status: {on_track|at_risk|delayed}
---
*Regular progress update. Work continues as planned.*
Interactive Prompts (if --interactive):
Q: What percentage complete is this task? [0-100]: 60
Q: What have you accomplished since the last update?: Implemented API endpoints, added unit tests
Q: Are there any blockers or challenges? [y/n]: n
Q: When do you expect to complete this? [date or duration]: 2 days
Auto-Fill Fields:
## Request for Review
**Status**: Awaiting Feedback
**Requested by**: {git_user_name}
**Request date**: {current_timestamp}
**Priority**: {auto_determine_priority}
## Context
{extract_from_recent_work}
## Work Completed So Far
{summary_of_commits_and_artifacts}
### Deliverables Ready for Review
{find_completed_artifacts}
- [ ] {deliverable_1} - `{path}` - {description}
## Feedback Requested On
### Critical Decisions
{if_applicable}
1. {extract_from_TODO_comments_or_ADRs}
### Areas Needing Validation
{if_applicable}
- {area_from_FIXME_or_comments}
## Specific Questions
{if_provided_via_interactive_or_content}
1. {question_1}
2. {question_2}
## Timeline Impact
{if_applicable}
- Waiting since: {current_date}
- Impact if delayed: {estimate_based_on_dependencies}
## How to Provide Feedback
1. Review the deliverables listed above
2. Comment on this issue with your feedback
3. Indicate approval or request changes
---
*Awaiting your review to proceed.*
Interactive Prompts (if --interactive):
Q: What priority level? [low/medium/high/urgent]: high
Q: What specific areas need feedback?: Architecture decision for data storage
Q: What questions do you have?: Should we use PostgreSQL or MongoDB?
Q: What deliverables are ready?: .aiwg/architecture/sad.md, .aiwg/architecture/adr-001.md
Q: Are there any blockers while waiting? [y/n]: n
Auto-Fill Fields:
## Blocker Alert
**Status**: Blocked
**Reported by**: {git_user_name}
**Reported date**: {current_timestamp}
**Severity**: {critical|high|medium|low}
**Impact**: {project-wide|phase|task|minimal}
## Blocker Description
{blocker_description_from_user_or_interactive}
## Root Cause
{if_known}
{description}
{if_unknown}
Under investigation. Initial observations:
- {observation_from_logs_or_errors}
## Impact Assessment
### Immediate Impact
- Work stopped on: {current_branch_or_task}
- Affects: {related_issues_or_components}
### Downstream Impact
{check_for_dependent_issues}
- Blocks: #{issue_numbers_from_dependencies}
- Timeline impact: {estimate_delay}
### Risk Level
- **Technical Risk**: {assess_based_on_severity}
- **Schedule Risk**: {assess_based_on_timeline}
- **Quality Risk**: {assess_based_on_impact}
## Proposed Solutions
### Option 1: {solution_name}
{from_user_or_interactive}
- **Description**: {description}
- **Effort**: {estimate}
- **Risk**: {assessment}
## Recommended Action
{recommendedaffectedwork} - Issue: #{issue
Interactive Prompts (if --interactive):
Q: Describe the blocker: External API rate limit reached, blocking integration tests
Q: Severity? [critical/high/medium/low]: high
Q: Impact scope? [project-wide/phase/task/minimal]: phase
Q: Root cause known? [y/n]: y
Q: Root cause description: Third-party API limits to 100 req/hour, need 500+
Q: Proposed solution: Use paid tier or implement request queuing
Q: Recommended action: Upgrade to paid tier ($99/mo)
Q: Does this block other issues? [issue numbers or n]: 18, 19
If --content provided:
{content_exactly_as_provided}
If interactive:
Q: Enter your comment (multiline, end with Ctrl+D):
{user_types_content}
Show Preview:
## Comment Preview
Issue: #{issue_number} - {title}
Type: {comment_type}
---
{rendered_template}
---
Post this comment? [y/n]:
Auto-post if non-interactive:
Skip preview and post immediately unless --dry-run.
GitHub:
# Post comment
gh issue comment {issue_number} --body "{comment_body}"
# Optionally add label
gh issue edit {issue_number} --add-label "{label_based_on_type}"
Gitea:
# Post comment using MCP
mcp__gitea__create_issue_comment \
--owner {owner} \
--repo {repo} \
--issue_number {number} \
--body "{comment_body}"
# Optionally update issue state or labels
Label Mapping:
| Comment Type | Suggested Label |
|---|---|
progress | in-progress |
feedback | needs-review |
blocker | blocked |
custom | None |
If blocker notification:
# Find dependent issues
gh issue list --search "depends on #{issue_number}" --json number
# Notify dependent issues
for dep_issue in ${dependent_issues}; do
gh issue comment ${dep_issue} --body "⚠️ Dependency blocked: Issue #{issue_number} is currently blocked. See #{issue_number} for details."
done
If feedback request:
# Assign reviewers if specified
gh issue edit {issue_number} --add-assignee {reviewers}
# Add to project board "Needs Review" column (if applicable)
gh project item-add --owner {owner} --project {project} --issue {issue_number}
## Comment Posted
**Issue**: #{issue_number} - {title}
**Type**: {comment_type}
**Posted by**: {user}
**Posted at**: {timestamp}
### Summary
{brief_summary_of_comment}
### Actions Taken
- [x] Comment posted to issue #{issue_number}
- [x] Label added: {label}
{if_applicable}
- [x] Dependent issues notified: #{numbers}
- [x] Reviewers assigned: @{usernames}
### Next Steps
{based_on_comment_type}
- For progress: Continue work, post next update in {timeframe}
- For feedback: Await review from {reviewers}
- For blocker: {recommended_action}
### Links
- Issue: {issue_url}
- Comment: {comment_url}
---
Comment successfully posted.
When to use:
What to include:
Auto-detection:
When to use:
What to include:
Auto-detection:
When to use:
What to include:
Auto-detection:
When to use:
What to include:
.aiwg/config.yamlissue_comment:
default_type: progress
auto_add_labels: true
label_mapping:
progress: "in-progress"
feedback: "needs-review"
blocker: "blocked"
notify_dependencies: true
auto_assign_reviewers: false
templates:
progress: "templates/issue-comments/progress-update.md"
feedback: "templates/issue-comments/feedback-needed.md"
blocker: "templates/issue-comments/blocker-found.md"
Frequency:
Content:
✅ What's done (concrete deliverables)
🔄 What's in progress (current focus)
📅 What's next (upcoming work)
⚠️ Risks or blockers (if any)
Example:
## Progress Update
Week of Jan 8-12, 2025
✅ Completed:
- Implemented issue sync detection logic
- Added 15 unit tests (coverage now 85%)
- Created issue comment templates
🔄 In Progress:
- Writing integration tests
- Documenting CLI usage
📅 Next:
- E2E testing with live repos
- Update README and docs
⚠️ No blockers. On track for Jan 15 completion.
Be Specific:
❌ Bad: "Please review this"
✅ Good: "Please review the architectural decision in .aiwg/architecture/adr-017.md. Specifically, should we use polling or webhooks for issue updates?"
Provide Context:
Context: We need real-time issue updates but GitHub has rate limits.
Options:
1. Polling every 5 minutes (simple, might hit rate limits)
2. Webhooks (complex setup, requires server)
Recommendation: Start with polling, add webhooks later if needed.
Question: Does this trade-off make sense for v1?
Urgency Levels:
| Severity | Response Time | Escalation |
|---|---|---|
| Critical | Immediate | Escalate after 1 hour |
| High | Same day | Escalate after 1 day |
| Medium | 1-2 days | Escalate after 1 week |
| Low | Best effort | No escalation |
Example:
## Blocker: GitHub API Rate Limit
Severity: High
Impact: Phase-level (blocks all integration tests)
Root Cause: Exceeded 5000 req/hour limit during test runs.
Proposed Solutions:
1. Use authenticated requests (10x higher limit) - 2 hours effort
2. Implement request caching - 4 hours effort
3. Reduce test frequency - 30 min effort (temporary)
Recommendation: #1 (authenticated requests) for permanent fix, #3 as immediate workaround.
Work Stopped:
- Integration test development
- CI pipeline testing
Escalate if: Not resolved by EOD (blocks sprint completion)
Project-Specific Templates:
Create custom templates in .aiwg/templates/issue-comments/:
# custom-deployment-update.md
## Deployment Progress
**Environment**: {environment}
**Release**: {version}
**Status**: {status}
### Deployment Steps
- [x] Build completed
- [x] Tests passing
- [ ] Deployed to staging
- [ ] Smoke tests passed
- [ ] Deployed to production
### Rollback Plan
{rollback_instructions}
Use with:
/issue-comment 17 --template custom-deployment-update
Automatically @-mention relevant people:
## Feedback Needed
@architect - Please review the database schema design
@security-team - Please validate the auth approach
@product-manager - Please confirm this meets requirements
{rest_of_feedback_request}
Configure in .aiwg/config.yaml:
issue_comment:
auto_mention:
feedback:
- "@architect"
- "@security-team"
blocker:
- "@tech-lead"
- "@project-manager"
Include Diagrams:
## Architecture Update
Current flow:
\`\`\`mermaid
graph LR
A[Client] --> B[API]
B --> C[Database]
\`\`\`
{rest_of_comment}
Include Code Snippets:
## Implementation Update
Added authentication middleware:
\`\`\`typescript
export const authMiddleware = async (req, res, next) => {
const token = req.headers.authorization;
// ... implementation
};
\`\`\`
{rest_of_comment}
gh issue commentThis command succeeds when: