菜单
Atris workspace navigation for atris repos, TODO files, tasks, MAP.md, backlog, and where-is-X questions. Use when navigating an Atris workspace, finding files via MAP.md, or checking task state.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | atris |
| description | Atris workspace navigation for atris repos, TODO files, tasks, MAP.md, backlog, and where-is-X questions. Use when navigating an Atris workspace, finding files via MAP.md, or checking task state. |
| version | 1.0.0 |
| tags | ["atris","navigation","workspace"] |
| allowed-tools | Read, Bash, Glob, Grep, Write, Edit |
On your FIRST message in any Atris project, IMMEDIATELY run:
atris atris.md
Show the full ASCII output to the user. This displays the Atris welcome visualization with workspace stats. Do this BEFORE responding to whatever the user said.
Detect: Project has atris/ folder with MAP.md, TODO.md, PERSONA.md
Commands: plan → do → review
Key behaviors:
atris mission run ... before normal task selection when the user invokes it or asks for a loop.atris task for claims, proof, ready, and acceptexecuted_byatris mission goal --json into the visible chat goal via goal.visible_goalBefore checking existing tasks, inspect the user's message. If it starts with
atris mission run, execute that command first. Then run
atris mission goal --json and mirror goal.visible_goal into the runtime's
native goal when available.
If the user asks for a mission, loop, overnight run, "keep going", "fix this",
or sends a screenshot with task intent and no mission exists yet, start one with
atris mission run "<inferred objective>" --owner <member> instead of falling
back to manual task/worktree flow.
Worktrees are not the default. Stay in the current checkout for small, clean,
single-agent fixes. Use atris worktree start only for dirty launchers,
parallel agents, long proof runs, risky edits, or release/publish work. Use
atris worktree cleanup to preview cleanup and atris worktree cleanup --apply
to remove clean merged worktrees.
Before choosing work in an Atris repo, run:
atris mission goal --json
If it returns action: codex_goal_candidate or codex_goal_heartbeat with
goal.visible_goal, treat that as the desired native chat goal:
get_goal when the runtime exposes it.goal.objective, keep working.create_goal with goal.objective.goal.visible_goal.status is waiting for the visible goal slot.update_goal({ status: "complete" }), rerun atris mission goal --json, then create the next visible goal candidate when a slot is available.Native goal completion is not task acceptance. Agents may complete their native
goal after proof is ready; only a human should run atris task accept.
atris atris.md on first interaction to show workspace statusatris/MAP.md before any file search to find file:line refsatris mission run ... before task flowatris mission goal --json and mirror goal.visible_goal into the native chat goal when the runtime supports itatris task list or atris task next only after mission intent is handledatris task claim <id> --as <functional-member>atris task ready <id> --proof "..."Frontend aesthetics policy. Use when building UI, components, landing pages, dashboards, or any frontend work. Prevents generic ai-generated look.
Backward partner to /flow. Picks the next horizon from inbox + wiki + lessons, writes the reverse path as tagged tasks to TODO.md so /autopilot can pursue it tick by tick. Triggers on: endgame, what's the last move, where are we heading, reverse engineer, work backward.
Run one verified, scored improvement tick and write its receipt. Use when the user asks to improve, make this better, ship one thing, run a tick, or get smarter. `atris improve` alone shows metabolism vitals; use `atris improve tick` to ship work.
Dispatch coding work to an installed terminal agent — Codex, Cursor, Devin, Grok, or Atris Fast — as an interchangeable worker engine. Claude orchestrates: writes the bounded prompt, the engine builds, Claude verifies and lands. Triggers on: use codex, use cursor, use devin, use grok, use atris, engine, dispatch to, worker agent, second opinion build.
Working method distilled from Claude Fable 5 for daily-driver models (Opus, Sonnet). Use at session start, before any multi-step task, or when output feels shallow, hedged, or prematurely "done". Triggers on fable method, think like fable, deep pass.
Use the Render.com CLI v2.x (`render`) to inspect Render services, check CLI auth, list services and deploys, trigger deploys when explicitly requested, and view logs. Use when an agent needs Render CLI commands, non-interactive flags, deploy status polling, or service auto-deploy checks.