convez.ai
SkillProdutividadeIniciante
rico2035

Status — TLDR de Sessão Claude Code

rico2035/claude-skill-status

TLDR automático: shipped, committed, pending, next — em um screen

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill que fornece um snapshot terse do estado da sessão em 15 segundos: o que foi entregue, o que está commitado, o que está pendente e próximos passos. Lê logs git e estado do projeto para gerar um relatório estruturado e sem fluff.

Instalação

$npx convez-ai rico2035/claude-skill-status

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
57%
Último commit
há 5 meses
Versão
v2.0.0
Issues abertas
0
Contribuidores
1

Casos de uso

  • Retomar contexto após pausa ou context-switch em sessão de código
  • Handoff rápido de código entre desenvolvedores
  • Checkpoint visual de progresso em sessões longas de development

Veredito

Ferramenta minimalista e bem-executada para quem trabalha com Claude Code. Resolve um problema real (perda de contexto em sessões longas) com formato padronizado e legível. Manutenção ativa, instalação trivial.

README

A tradução para o português deste README ainda está na fila.

/status — Claude Code Session TLDR Skill

Give your Claude Code session a TLDR snapshot in 15 seconds: what shipped, what's committed, what's pending, what's next.

Built by Ric S Kolluri at Novatar.ai. MIT-licensed. Attribution required (see License).


What it does

You type /status (or "give me an update", "where are we", "what did we ship") and Claude returns a terse, structured TLDR:

## TL;DR

### Shipped
- R046 Customer Intelligence Brief scaffold (skill + template + intake checklist + PRIME docs)
- 7 community skills installed globally, all malware-scanned clean

### Committed (this session)
| Hash | Subject |
|---|---|
| `10123a3` | feat(prime): customer intelligence brief scaffold (R046, REQ-123) |
| `82bbe73` | chore(prime): close R046 — mark released, move folder |

### Pending (your action)
1. Commit the REQ-129 entry in REQUEST-LOG.json (parallel session left it uncommitted)
2. Decide whether to schedule the broken-cloak-gate cleanup as a new REQ

### Next step
You commit the REQUEST-LOG.json housekeeping, then we can start a new task.

If the session is genuinely done, you get exactly this and nothing else:

Nothing pending. Ready for new task.

What it does NOT do

  • No file writes. This is a read-only TLDR generator. (If you want state saved to disk for resumption, use /checkpoint instead.)
  • No long prose. If you want a structured prose recap, use /recap instead.
  • No project-level status. Session-scope only. For project state, use /roadmap.
  • No follow-up questions. It's a snapshot, not a conversation.
  • No "everything looks good" fluff. Either there's a Next step or it hits the terminal state. Pick one.

Why this exists

Long coding sessions in Claude Code accumulate context fast. When you context-switch, return after a break, or hand off to a teammate, you need to know in seconds: did the session ship anything, what's committed, what's still pending. A consistent, terse output format makes that instant.

/status enforces the format so every snapshot looks the same: commit hashes in backticks, file paths as clickable markdown links, no em dashes, no AI vocabulary, no preamble. You can read it in one screen.


Install — single command

macOS / Linux / WSL

curl -fsSL https://raw.githubusercontent.com/rico2035/claude-skill-status/main/install.sh | bash

Windows (PowerShell)

irm https://raw.githubusercontent.com/rico2035/claude-skill-status/main/install.ps1 | iex

Inspect before piping. If you prefer to read the installer first (recommended for any curl | bash pattern), clone the repo and run locally:

git clone https://github.com/rico2035/claude-skill-status.git
cd claude-skill-status
bash install.sh        # or: powershell ./install.ps1

The installer copies status/SKILL.md into ~/.claude/skills/status/SKILL.md. Once installed, /status is available in every Claude Code session.


Usage

In any Claude Code session:

/status

That's it. Claude reads the current TodoWrite state, runs git log and git status --short, and formats the TLDR.

You can also trigger it without the slash by saying any of:

  • "give me an update"
  • "what's the status"
  • "where are we"
  • "what did we ship"
  • "TLDR what just happened"
  • "TL;DR"

Triggering it after a break (e.g., starting a new session in the same project) is exactly when it pays off.


Dependencies

WhatRequired?Why
Claude CodeYesThis is a Claude Code skill
gitYes (for git log / git status)The skill enumerates commits and uncommitted changes
bash (macOS/Linux/WSL) or PowerShell (Windows)Yes (for the installer only)Pre-installed on every modern system

That's the entire dependency surface. No sub-skills, no MCP servers, no external CLIs.


Companion skills (not required, but related)

If you like /status, you may also want:

SkillWhen to useWhere it lives
/recapLong-form structured prose recap, not a TLDR(your own or a community skill)
/checkpointPersist session state to disk so you can resume across sessions(your own or a community skill)
/roadmapProject-level status (not session-level)(your own or a community skill)

/status deliberately stays in its lane: read-only, session-scoped, snapshot only.


Configuration

None. The skill is intentionally configuration-free. The output format is enforced for consistency — every TLDR looks the same, every session.

If you want a different format, fork the repo and edit status/SKILL.md. The whole skill is a single markdown file.


License

This project is licensed under the MIT License with one explicit attribution expectation:

You must mention "Built by Ric S Kolluri at Novatar.ai" in any derivative skill or product that uses this work.

The MIT license already requires the copyright notice to be retained in all copies and substantial portions. This README is restating that requirement in plain language so it's not missed. Concretely:

  • If you fork this repo and ship a modified version, keep the LICENSE file and credit the original work.
  • If you redistribute the skill bundled inside another product, include the attribution in your own credits/about page.

The skill is otherwise free to use, fork, modify, and redistribute under MIT terms. See LICENSE for the full legal text.


Security

No API keys, tokens, credentials, or private endpoints ship with this repo. The skill is plain markdown — no code, no network calls, no background processes. See SECURITY.md for the full trust model and how to report security issues.


Contributing

Issues and pull requests welcome. For substantive changes to the output format, open an issue first to discuss — the format is intentionally rigid because consistency is the whole value proposition.

When contributing:

  • Keep the "no AI tells" writing style: no em dashes, no AI vocabulary (leverage, robust, seamless, harness, transformative, etc.), short sentences, numbers lead.
  • Don't add network calls, file writes, or telemetry. /status stays read-only and local.
  • Keep the skill a single file. If it needs supporting files, reconsider whether the addition belongs in /status or a separate skill.

Contact

Ric S Kolluri — Novatar.ai

For security issues, see SECURITY.md. For everything else, open a GitHub issue or email above.


Companion repo

If you build agents and need an 18-month strategic scaling report generator, check out /cscale — a much heavier orchestrator skill from the same author.

Ver o original no GitHub

Bom para

Relacionados