菜单
iterate with spawned agents until stable. use for verification with BOUNDED scope and explicit exit criteria. NOT for generating opinions to reconcile.
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
Interact with local Chrome browser session (only on explicit user approval after being asked to inspect, debug, or interact with a page open in Chrome)
orchestrate multiple amp agents with bidirectional tmux communication. use for multi-hour autonomous runs with INDEPENDENT parallel tasks. NOT for review courts or generating opinions to reconcile.
systematic investigation and debugging methodology. hypothesis-driven analysis with verification agents. use for incident response, codebase archaeology, dependency mapping, root cause analysis, or any investigation requiring verified findings.
apply documentation philosophy — explain why, not what. use for jsdocs, READMEs, inline comments.
git workflows for agents: ship (stage → commit → push), worktree (parallel branches), hunks (selective staging). never force push, never git add -A, conventional commits. triggers on: commit, push, stage, ship, git add, worktree, hunks, selective staging.
structured investigation-to-PR workflow for linear issues. hypothesis-driven debugging with browser validation and counterfactual proof. use when investigating a linear ticket, fixing a UI bug, or shipping a fix with evidence. triggers on: investigate issue, fix bug, nexus-fix, AXM-, linear issue.
| name | rounds |
| description | iterate with spawned agents until stable. use for verification with BOUNDED scope and explicit exit criteria. NOT for generating opinions to reconcile. |
iterate with spawned agents until results stabilize.
prerequisite skills: spawn, coordinate, report
before running multi-round review, ask:
rounds is for verification where first-clean can't be trusted. don't use it to generate opinions for reconciliation.
round 5 finding issues after round 4 was clean is common. run 2-3 verification rounds minimum after issues stop appearing.
round N:
1. spawn agents for task (parallel if independent)
2. wait for reports
3. if issues → spawn fix agents → wait → proceed to N+1
4. if clean → proceed to N+1 anyway (verify the clean)
5. repeat until 2+ consecutive clean rounds
tell spawned agents which foundation skills to load. example for code review:
load the review skill before evaluating. load the write skill for report quality.
adapt based on task:
review, writereview, digreview, write, documentwhen agents find issues requiring product decisions ("remove feature X or implement it?"), pause and ask user. don't make product decisions unilaterally.
rounds can orchestrate parallel spar debates. see spar skill for the full composition protocol, interface contract, and termination conditions.
after completion:
| round | result |
|---|---|
| 1 | 3 issues → fixed |
| 2 | clean |
| 3 | clean |
list all changes made across fix phases.