focused-fix
CommandDeep-dive feature repair — systematically fix an entire feature/module across all its files and dependencies. Usage: /focused-fix…
/cs-inbox-triage — Recurring email triage execution. Reads 7-file KB built by /cs-inbox-setup. Classifies recent emails, drafts replies (NEVER SENDS), delivers report, updates KB. Run 1-3x/day or on demand. Halts with clear message if KB missing.
$ npx -y skills add alirezarezvani/claude-skills --agent claude-code
disable-model-invocation: true name: "cs-inbox-triage" description: "/cs-inbox-triage — Recurring email triage execution. Reads 7-file KB built by /cs-inbox-setup. Classifies recent emails, drafts replies (NEVER SENDS), delivers report, updates KB. Run 1-3x/day or on demand. Halts with clear message if KB missing."
**Command:** `/cs-inbox-triage`
The `cs-inbox-triage` persona processes your inbox using the knowledge base built by `/cs-inbox-setup`. Designed for recurring runs (1-3x/day) with light intake — most invocations skip questions and run with KB-default preferences.
**Do NOT run if** the KB doesn't exist yet — the skill will halt and direct you to `/cs-inbox-setup` first.
> **This skill creates drafts. It NEVER sends.**
This is the safety property that makes the skill safe to run automatically. The `draft_safety_validator.py` enforces it post-run. Any send-shaped tool call in the action log fails validation.
If you want the skill to send for you: don't. Review the drafts in your email client and send them yourself. This is by design.
Most runs skip both questions entirely.
| Q | Asked when | Default if skipped | |---|---|---| | Q1 — Override default 9h window? | On-demand run outside normal cadence | use cadence default | | Q2 — Skip categories this run? | User invocation includes skip-intent ("skip newsletters") | run all categories |
After reading the KB:
1. **Determine search window** — cadence + now → window_start (default 9h for 2x/day; overlap prevents missed emails) 2. **Search email provider** — primary (inbox + sent after window_start) + secondary (starred unread) 3. **Classify** — apply taxonomy; skip lowest-priority threads (newsletters/automation) without reading 4. **Research new senders** — web search for opportunity senders not in tracker/blocklist 5. **Generate recommendations** — apply `evaluation-framework.md` if exists; categorize TAKE IT / WORTH CONSIDERING / PASS / FLAG FOR REVIEW 6. **Draft replies** — match voice from `email-patterns.md`. NEVER SEND. 7. **Deliver report** — honor `email-taxonomy.md` report preferences (email / file / chat) 8. **Update KB** — append new declines to `blocklist.md`; update `tracker.md` with new/resolved follow-ups 9. **Internal log** — write `triage-log/<YYYY-MM-DD>-<run-label>.md` 10. **Empty inbox handling** — still produces minimal report; flags overdue tracker items
# 1. Pre-flight — read + validate KB (fail-fast if missing)
python ${CLAUDE_PLUGIN_ROOT}/skills/inbox-triage/scripts/kb_reader.py --workspace ${WORKSPACE}
# 2. Compute search window
python ${CLAUDE_PLUGIN_ROOT}/skills/inbox-triage/scripts/search_window_calculator.py \
--cadence 2x-daily --now $(date -u +%Y-%m-%dT%H:%M)
# 3. Execute Steps 2-10 (described in SKILL.md). For each step, log to:
# ${WORKSPACE}/Email/triage-log/<date>-<label>.md
# 4. Post-flight — verify NEVER-SEND held
python ${CLAUDE_PLUGIN_ROOT}/skills/inbox-triage/scripts/draft_safety_validator.py \
--action-log ${WORKSPACE}/Email/triage-log/$(date +%Y-%m-%d)-*.md
# Failure here is critical — halt + alert user immediately1. **DRAFTS ONLY — NEVER SEND.** Non-negotiable. 2. **Fail-fast on missing KB.** Halt cleanly; direct to setup. 3. **Honor the KB.** Documented preferences are source of truth; don't override with judgment. 4. **Privacy.** No credentials in KB; reference threads by ID for sensitive content. 5. **Transparency.** Note every KB change in the triage log. 6. **First runs need oversight.** Documented — system learns from your edits and overrides.
| Category | When | Output | |---|---|---| | **TAKE IT** | Meets criteria from `evaluation-framework.md` | Recommend engaging; draft reply | | **WORTH CONSIDERING** | Has potential, needs user judgment | Surface key context; draft reply for user to edit | | **PASS** | Doesn't meet criteria | Brief "why"; draft polite decline | | **FLAG FOR REVIEW** | Unusual; needs direct user decision | Surface fully; NO draft (user decides response shape) |
---
**Version:** 1.0.0 **Source:** Path-B direct conversion of `megaprompts/07-inbox-triage-megaprompt.md`
411 production-ready skills for Claude Code and 12 other coding agents, covering engineering, product, marketing, compliance, finance and the C-suite. We test whether each skill loads when you ask in plain words, and publish the results.
Deep-dive feature repair — systematically fix an entire feature/module across all its files and dependencies. Usage: /focused-fix…
Clean up merged branches locally and on remote, keeping only main, dev, and gh-pages.
Stage working tree changes and create a Conventional Commit (no push).
Stage, commit, and push the current branch following git governance rules.
Comprehensive audit pipeline for skills, plugins, agents, and commands. Validates structure, quality, security, marketplace…
Context: runloop resumes a devbox by re-running the same entrypoint script, which invokes `cn serve --id <agentId> ...`. Because…
Shortcut for "continue working" — pick the work in flight back up exactly where it stopped and keep going until it is done or…
Spawn a read-only sidecar subagent to investigate something in this checkout without editing files, moving branches, or touching…