mit einem Klick
awaken
新仓库落地时强制执行发现仪式,禁止模板复制
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Menü
新仓库落地时强制执行发现仪式,禁止模板复制
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Evidence-gated completion check. Use before committing, creating PRs, or claiming work is done. Pairs with verification-spec at task start.
Systematic knowledge extraction from open-source projects. Use when: user shares a repo/link to study, says 偷师/steal/学习/研究, or wants to analyze another project's patterns for adoption.
Launch a real interactive claude CLI session in a target directory (typically a worktree) with a pre-seeded prompt. Use when the current session cannot absorb the task context — e.g., N parallel worktree jobs that would blow the context window, or when you need the user to watch progress in a visible terminal. Each spawned session runs independently in its own Windows Terminal tab.
FileNotFound recovery protocol. Before reporting a missing-file error, suggest near-misses by basename similarity (difflib threshold 0.4) so typos and stale paths fail loudly with actionable hints.
Paste-by-reference guard for Edit calls. When new_string would echo >30 lines of existing file content, use {{file:path:start:end}} reference instead and expand before submission.
File-IO protocol for parent ↔ subagent communication. Parent writes input.txt + context.json, subagent appends to output.txt with [ROUND END] sentinels. Used with Monitor tool for live observation.
| name | awaken |
| description | 新仓库落地时强制执行发现仪式,禁止模板复制 |
| origin | R81 loki-skills-cli steal — forced re-discovery pattern |
| source_version | 2026-04-18 |
You are landing in an unfamiliar repository. Before writing a single line of code, you must complete this ritual in full. The goal is understanding, not setup — you must be able to explain this repo from memory before touching it.
BANNED: copying any section verbatim from README, CLAUDE.md, or any template. Write each section from understanding, not from paste.
Run the following to read project fundamentals:
cat README.md 2>/dev/null; cat CLAUDE.md 2>/dev/null; ls docs/ 2>/dev/null
Gate: Before continuing to Step 2, you must be able to state the project's primary purpose in one sentence — from memory, not from re-reading. If you cannot, re-read and try again until you can.
Spawn one sub-agent with the following literal-path arguments (no variables, no relative paths):
SOURCE_DIR="<absolute-path-of-new-repo>"
DEST_DIR="<absolute-path-of-new-repo>/.remember/"
Sub-agent task: search for architecture decision records and key design documents:
grep -r "ADR\|architecture\|AGENTS\|spec" $SOURCE_DIR --include="*.md" -l
Sub-agent writes findings to $DEST_DIR/onboard-trace-MMDD.md (MMDD = today's date, e.g., 0418).
Write .orchestrator-instance.md in the project root. This file must contain:
Gate: Do not write this file until you can explain each convention from memory without re-reading the source files. Write from understanding.
Commit the instance file on a dedicated onboarding branch:
git checkout -b onboard/<repo-slug>
git add .orchestrator-instance.md
git commit -m "chore(onboard): Orchestrator awakens in <repo-slug>"
Replace <repo-slug> with the repo directory name (e.g., my-project).
Append a one-paragraph retrospective to .remember/retrospectives.md noting what surprised you about this repo's conventions. If .remember/ does not exist, create it first.
Format:
## <repo-slug> — YYYY-MM-DD
<One paragraph. What was unexpected? What did you assume that turned out to be wrong? What convention required the most re-reading to understand?>