t3

Periodic holistic architectural review — the third of teatree's three review tiers (design-time `architecture-design`, per-PR deterministic `check_antipatterns.py`, periodic holistic `ac-reviewing-codebase`). Walks the whole tree for judgement-tier anti-patterns and BLUEPRINT.md staleness that no single diff can catch, implements what it finds, and pushes one PR. Dispatched automatically by `ArchitecturalReviewScanner` on a time or merge-count cadence — not user-invoked.

Draft a reply to an inbound question, DM the user for approval, post on confirmation. Use when the loop routes a `question` intent to phase `answering`, or when asked to answer a Slack/GitLab thread on the user's behalf.

Architecture pre-check companion. Loaded transitively by implementation skills (code, ticket-for-features, retro-for-skill-changes) to force an architecture pass — BLUEPRINT alignment, FSM phase boundaries, extension-point contracts, component boundaries, dependency direction, test surface, resilience invariants, removability — BEFORE any code is written.

The check-in surface — a SHORT "what did I miss" report, the session task/TODO lists, the pending deferred questions, and the daily follow-up routine (new tickets, ticket statuses, PR reminders). Use when the user says "what did I miss", "checking", "catch me up", "what changed since", "ask me the questions", "task list", "todos", "what's on my list", "my tasks", "followup", "check status", "advance tickets", or "PR reminders".

Writing code with TDD methodology. Use when user says "implement", "write", "add", "code", "feature", "build", or is actively coding a solution.

Push retro improvements to a branch, open a PR, and optionally create upstream issues. Use when user says "contribute", "push improvements", "push skills", or after retro creates a local commit.

Troubleshooting and fixing — something is broken, find and fix it. Use when user says "broken", "error", "not working", "crash", "blank page", "can't connect", "debug", or reports any failure.

Submit a plain-English directive about how teatree itself should behave — captured verbatim, interpreted into a typed mechanism sketch, human-ratified via Slack/questions, then implemented through the gated pipeline. Use when the user says "/t3:directive ...", "directive:", "from now on teatree should...", "teatree must always/never...", or asks to change teatree's own behavior in natural language.

Dogfooding teatree's own CLI, loop, and statusline — two modes sharing one mechanics section for reading a tick and the rendered statusline. "Verify a change" is the run-it-yourself checklist applied after modifying CLI/loop/statusline code, before declaring it done. "Hunt for bugs" is proactive self-QA — dogfood the deployed loop, find/dedupe/confirm real bugs, file them, then fix them in worktrees. Use when the user says "dogfood", "dogfooding", makes a CLI/loop/statusline/scanner change, or says "bug hunt", "self-qa", or "hunt bugs".

Runs the idle-time "dreaming" memory-consolidation pipeline end to end with one command — replay recent transcripts + curated memories, distil drift into the ConsolidatedMemory ledger, cross-link / re-index / decay the memory files, run the §4 acceptance gates, triage each row into keep-as-memory vs core-gap → drive each core gap to a MERGED fix under the standing umbrella issue, and promote/stage eval candidates. Use when user says "dream", "dreaming", "consolidate memory", "run the dream pass", "memory consolidation", or wants the full dream pipeline.

End-to-end testing with Playwright — writing tests, running them, visual snapshots, writing the test plan into the e2e repo, and the pre-push visual QA gate. Use when user says "e2e", "playwright", "write e2e", "run e2e", "visual qa", "screenshot", "test plan", "post test plan", "post evidence", or is working with Playwright-based tests.

Reviewer-side quality gate for Playwright end-to-end specs. Load when reviewing a new or changed E2E test, deciding whether a spec is ready to land, or adopting an outside Playwright suite. Judges specs against Playwright's published best practices — user-visible behaviour over implementation, resilient role/label/test-id locators, web-first auto-retrying assertions instead of hard waits, per-test isolation, page-object structure, and runnable evidence — and tells the implementer what to fix before approval.

Use when the user wants to hand all current work from one Claude session to another (or to a not-yet-existing session) with a single command, or to transfer an in-flight TeaTree task from Claude to another runtime, or asks whether it is time to switch because Claude usage is getting high.

Read and act on the global operational-health chip — the green/yellow/red factory-health verdict and its known-issues registry. Use when the statusline health chip is yellow/red, or the user asks "what's wrong", "why is health red", "is the factory healthy", "known issues".

ENGAGES TEATREE FOR THE SESSION, and holds the standing rule that no work-bearing state is terminal. Loading this skill — or any skill declaring `requires: interactive` — writes the `.teatree-active` marker, one of the two conditions in `_loop_auto_load_active()` that arm the loop and statusline (#256); a session that never loads it stays unengaged, by design. Also holds teatree's Claude Code harness wiring: how skills are selected, how plugin hooks are registered, and which output belongs to the headless pipeline. Load it when ending an interactive session, when a session-end report names stranded work, or when deciding what to do with uncommitted, unpushed, untracked or unmerged work. Teatree's own architecture and coding rules are `/t3:internals`; the dogfooding procedure is `/t3:dogfooding`.

How teatree is BUILT and how to change it safely — architecture, lifecycle phases, key models, the overlay API, the `t3` CLI reference, and the management-command rules whose violation fails SILENTLY (a `typer.Exit` under `call_command` exits 0, so CI reports green on a real failure). Load it when writing or reviewing teatree's own code, or when building an overlay on it. Carries no Claude Code harness wiring — that is `/t3:interactive` — and no dogfooding procedure — that is `/t3:dogfooding`.

The operating mode — one of five named presets (present / away / maintenance / low-token / off) deciding which loops run. Use when switching mode for a holiday or an unattended run, configuring the weekly schedule, answering the deferred-question backlog, or debugging the mode resolver.

Wrap up the current session — retro, structured result, pipeline handoff.

Platform-specific API recipes for GitLab, GitHub, Slack, and X (Twitter). Auto-loaded as a dependency by skills that interact with these platforms.

Trigger and manage reusable prompts — list the prompts in the DB, render one by name with its templated params, and point to the admin for authoring + version history. Use when the user says "prompts", "run a prompt", "trigger a prompt", "render a prompt", "list prompts", or "prompt library".

Conversation retrospective and skill improvement. Use when user says "retro", "retrospective", "lessons learned", "improve skills", "what went wrong", "auto-improve", or at the end of a non-trivial work session.

Code review — self-review before finalization, giving review, receiving review feedback. Use when user says "review", "check the code", "feedback", "review comments", "quality check", or is in a review cycle.

Batch review requests — discover open PRs, validate metadata, check for duplicates, post to review channels. Use when user says "request review", "send for review", or wants to batch-notify reviewers.

Cross-cutting agent safety rules — the always-embedded core carrying every non-negotiable's trigger and verdict, each naming the skills/rules/references/ file with its full text. Auto-loaded as a dependency by other skills.

Single in-session entrypoint that auto-orchestrates the whole eval picture — model-free deterministic lanes (the eval-coverage gate `t3 eval coverage`, pinned-regressions) plus the transcript AI/trajectory lane (prepare → produce transcripts in-session → grade) — and prints one unified results table. Use when running the full eval suite, producing recorded transcripts, or deciding between `t3 eval run` (AI evals) and `t3 teatree run tests` (deterministic tests).

Scans today's TLDR AI and The Rundown AI editions for ideas that could improve teatree, fetches the full article for promising items, and hands each concrete t3-improvement candidate back through the result envelope's article_suggestions field. The loop queues each behind the ask-gate (PendingArticleSuggestion) for per-article user approval before any souliane/teatree issue is filed, and DMs the batch to the user. Use when user says "scan news", "scanning news", "ai newsletter scan", "improvement scan", "tldr scan", or runs the daily improvement routine.

Bootstrap and validate teatree for local use — prerequisites, config, skill symlinks, optional agent hooks, and Django project scaffolding. Use when user says "setup skills", "install skills", "bootstrap skills", or needs first-time teatree installation.

Delivery — committing, pushing, creating MR/PR, pipeline monitoring, review requests. Use when user says "commit", "push", "PR", "merge request", "pull request", "finalize", "deliver", "ship", or is in the delivery phase.

Rendering tables and formatting messages for Slack — the native Block Kit table block, the monospace fence fallback, and the mrkdwn gotchas (no pipe tables, single-asterisk bold, angle-bracket links). Auto-loaded as an overlay companion for work that posts to Slack.

Maintenance sweep across all your open PRs/PRs — merge the default branch, fix conflicts, monitor CI, push, and (per-repo policy) optionally squash-merge each PR before moving to the next. Never rebases. Use when user says "sweep PRs", "update all my PRs", "merge main into open PRs", or wants to keep open PRs up to date with main.

Use when sweeping stale, lost, or abandoned worktrees, branches, or stashes that are NOT actively being worked — deciding per item whether to salvage unmerged work to a fresh PR, delete a shipped/superseded/redundant item, push post-merge commits to a new PR, or keep an uncertain one. The judgment layer over `t3 <overlay> workspace emit` / `salvage` / `clean-all` (the mechanical reaper is `/t3:workspace`). Use when the user says "sweep the stale worktrees", "clean up lost/abandoned worktrees", "salvage or delete these branches", "what should I do with these stale worktrees/stashes", or triages the JSON from `workspace emit`.

Testing, QA, and CI — running tests, analyzing failures, quality checks, CI interaction, test plans, and posting testing evidence. Use when user says "run tests", "pytest", "lint", "CI failed", "pipeline", "test plan", "QA", or any test/CI task.

Ticket intake and kickoff — from zero to ready-to-code. Use when user says "I have ticket X", "new ticket", "start working on", "what should I do for this?", or provides a ticket/issue/PR link.

Review and act on the needs-triage assessor's queued recommendations — list PENDING PendingTriageRecommendation rows, approve or reject each, and on approval run `gh issue close/edit/comment` then stamp the row. Use when user says "triage issues", "triaging issues", "review triage recommendations", "assess needs-triage", or catches up on the triage-assessor batch.

WHEN to bring teatree core and registered overlays up to date with their default branch, and the safety guarantees of doing so. Use when the user wants to "sync teatree", "update teatree", "pull latest teatree", or after merging PRs that change framework/overlay code.

The bounded-WIP throughput dial — slow / medium / full / boost — plus the WRITE-parallel / MERGE-serial phase split and the per-ticket unattended delivery cycle. `boost` keeps `boost_concurrency = N` workers live; `full` arms a self-sustaining boost loop; `medium` (baseline) and `slow` cap concurrency. Use when the user says "wip", "go full speed", "blast the backlog", "boost", "parallel mode", "max throughput", "go wide", "slow down", "set wip", "work unattended", or "tackle tickets".

Environment and workspace lifecycle — worktree creation, setup, DB provisioning, dev servers, cleanup. Use when user says "create worktree", "setup", "start servers", "refresh DB", "cleanup", or any infrastructure task.