ClaudeMap

tstack

A set of everyday agent skills for shipping pull requests and multi-model review, packaged to run in both Claude Code and Codex.

Skills & Commandsskillsclaude-codeworkflowmulti-agent
Visit projectt3dotgg/tstack

tstack packages the skills its author runs every day for pull-request work: full-send takes a change all the way to a merged PR unattended (prove it, review it, file it, babysit it, merge it), full-stop stops at ready-to-merge, council and sol-loop loop review across providers until models agree, and file-pr, takeover, babysit-pr, modernize and leaving-pr-comments cover the surrounding PR chores. Skills call each other, install by copying into the Claude Code or Codex skills directories, and fall back to native tools when T3 Code features are unavailable.

What it does

  • End-to-end PR automation — full-send proves a change, reviews it, files the PR, babysits it through CI, and merges it with nobody watching; full-stop is the same pipeline that halts when the PR is ready for you to merge
  • Multi-model review loops — council collects review from three models on different providers and iterates until they agree; sol-loop runs a focused GPT Sol review loop
  • Composable skills — full-send calls sol-loop, file-pr and babysit-pr; takeover finishes someone else's PR, branch, or agent thread and keeps their credit; modernize rebases an old PR and judges whether it should still exist
  • Cross-agent by construction — one repo installs into ~/.claude/skills/ or ~/.codex/skills/, and skills use T3 Code features when available with fallbacks to native tools where they are not

When to use it

Solo maintainers who want an unattended path from local change to merged PR, and reviewers who want a second (and third) model to argue with before they approve. The skills assume a real GitHub workflow — an authenticated gh CLI and CI on the PR — rather than local-only tinkering.

Getting started

git clone https://github.com/t3dotgg/tstack
cp -R tstack/skills/* ~/.claude/skills/   # Claude Code
cp -R tstack/skills/* ~/.codex/skills/    # Codex

Things to know

council and sol-loop need subagents on models from other providers — T3 Code provides this with delegate_task, and otherwise you need a harness that can run those models. An authenticated GitHub CLI is required throughout, and babysit-pr's pr-wait script needs python3. The repo landed on 2026-10-24 and ships the author's personal daily setup — the README's own framing is "These are starting points. Edit them."

Repository

t3dotgg/tstack

Charted

Related on the map

  • SureForge

    Skill

    An instruction-only skill that makes agents research before they ask, plan before they build, and verify before they call anything done.

    • workflow
    • verification
    • quality
    • skill
    View

  • guizang-product-video-skill

    Skill

    Guizang's agent skill turns your repo's recent updates into a finished promo video — real components, code-rendered animation, an original code-written soundtrack — no video editor involved.

    • agent-skill
    • video
    • promo
    • marketing
    View

  • geo-sleuth

    Skill

    An agent skill that geolocates a photo with nothing readable in it — OSM geometry, elevation skylines and satellite tiles turn “where was this taken?” into a measured answer with an error radius.

    • agent-skill
    • geolocation
    • osint
    • openstreetmap
    View