メニュー
Use when repo-bug-finder discovery is complete and the user asks to package confirmed or probable findings into final Bug records, indexes, handoff materials, or an HTML report.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when asked to find Bugs, uncover more repository defects, audit one or many repos for security, reliability, data-integrity, recovery, permission, or architecture risks, or continue high-recall static Bug discovery.
Use when asked to review a skill's quality, test whether a skill works correctly, find why a skill behaves inconsistently or fails to trigger, check if a skill is ready to publish, harden a skill against known failure modes, or turn an observed failure into a repeatable test case.
Generate Outlook-compatible email templates (EML + HTML) through conversation. Three modes: Design (create from scratch), Import (replicate an existing .eml), Production (fill Excel data into a crystallized template). Use when user wants to: create or design an email template, generate an .eml file, make a newsletter, format an email for Outlook, import/replicate an email (导入/复刻邮件), design a 邮件模板, do 邮件排版 or 邮件设计, build pixel-perfect HTML email with Outlook compatibility. Triggers: weekly report email (周报邮件), product update email, event invitation (活动邀请邮件), announcement (公告邮件), company newsletter, .eml import, replicate email template, or make an email look professional/beautiful for Outlook. Handles visual design, EML generation, and EML import — not SMTP, sending, or account management. Without this skill, Outlook emails break because Outlook uses Word rendering which ignores modern CSS.
Vibe Deck — vibe-code professional slide presentations — describe what you want, AI builds it. Scaffolds a React + ECharts project, creates slides with charts, animations, theming, and PDF export. Use PROACTIVELY when the user mentions slides, deck, presentation, PPT, PPTX, slideshow, keynote, pitch deck, quarterly review, board meeting, investor update, sales deck, training deck, onboarding slides, report presentation, add a slide, build a deck, create slides, make a roadmap slide, put this data into a presentation, turn this Excel into slides, visualize this data as a deck. Also trigger when the user wants to modify, reorder, or delete slides in an existing slide-kit project. Also trigger when the user wants to share, export, or package the deck as a single HTML file for email or offline viewing. Chinese triggers: 做PPT, 做个deck, 写pptx, 创建演示, 制作幻灯片, 做幻灯片, 加一页, 新增slide, 做演示文稿, 工作汇报, 述职报告, 季度回顾, 方案展示, 写个汇报, 改一下这页, 调整幻灯片顺序, 删掉这页, 把数据做成图表展示, 帮我做个路线图, 导出单个HTML, 分享给别人看.
| name | repo-bug-finder:package |
| description | Use when repo-bug-finder discovery is complete and the user asks to package confirmed or probable findings into final Bug records, indexes, handoff materials, or an HTML report. |
Convert discovery readiness into a final handoff package. This subskill is packaging-only: it must not read new code facts, hunt for new Bugs, or promote scanner hits that were not triaged during discovery.
Input:
work/discovery-readiness.jsonwork/discovery-readiness.mdwork/architecture-signals.json and .md when presentOutput uses this standard package structure:
submit/
README.md
bug-audit-report.html
findings/P1/
findings/P2/
findings/P3/
findings/P4/
indexes/
knowledge/
quality/
standards/
Read references/packaging-workflow.md, references/package-output.md, references/bug-schema.md, references/risk-taxonomy.md, and references/interactive-html-report.md.
Export discovery readiness into the package structure:
python3 scripts/export_readiness_to_package.py <workspace> --project <name> --language zh
--workspace <workspace> is accepted as an agent-friendly alias for the positional workspace.
Convert only clean package-ready records into Bug Markdown files. When readiness provides summary.clean_package_ready_record_ids, export only those records; otherwise fall back to eligible confirmed/probable records from older readiness files.
Assign continuous IDs starting at BUG-0001 with no gaps.
Sort before assigning IDs: confidence high -> medium -> low, then priority P1 -> P2 -> P3 -> P4.
Put probable records in the final package only when the sole missing gate is trigger-path; label confidence: medium or low and state the trigger as a hypothesis.
Keep candidate, refuted, and merged records in candidate/quality coverage, not as final Bugs.
Generate indexes:
python3 scripts/generate_bug_index.py <workspace>/submit
python3 scripts/generate_candidate_index.py <workspace>/submit
Package scripts that operate on submit/ also accept --submit-root <workspace>/submit or --workspace <workspace>.
Run pre-package validation before HTML generation when final HTML is expected:
python3 scripts/validate_bug_package.py <workspace>/submit
Do not pass --repo-root by default. In this workflow, work/discovery-readiness.json is the package source of truth. Use --repo-root only when the user explicitly asks for source-path existence validation against local checkouts.
Generate the HTML report:
python3 scripts/generate_bug_report_html.py <workspace>/submit --language zh
Run final validation:
python3 scripts/validate_bug_package.py <workspace>/submit --require-html-report
Use --language en and English Bug headings only when the final deliverable is English.
work/architecture-signals.json.submit/findings/.submit/findings/ when readiness exposes clean package-ready IDs.Final package completion must produce:
submit/findings/P*/BUG-*.md for promoted confirmed/probable records.submit/indexes/findings.generated.json and submit/indexes/candidates.generated.json.work/package-validation/prepackage-validation.passed.json before HTML generation.submit/bug-audit-report.html when a final report is requested.validate_bug_package.py <workspace>/submit --require-html-report output when HTML is generated.The package structure, Bug schema, index format, and HTML report style are part of this subskill's own contract. Keep them stable for downstream readers.