Content
85%Weight 40%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A well-structured, actionable router skill with a clear phased workflow, guardrails, completion checklist, and excellent progressive disclosure into a large verified bundle. The main weakness is conciseness — the playbook table, asset-count table, and directory tree overlap, and the unimplemented CLI-prefix section adds little.
Suggestions
Remove or merge the redundant '数据资产规模' count table and '引用 / 跨链结构' directory tree with the Phase-4 playbook/methodology tables to cut overlapping content.
Drop the 'CLI 助记前缀' section or fold its naming convention into a single line, since the CLI is explicitly unimplemented and it adds no executable value.
Consolidate the repeated compliance red-lines (inline list vs. per-playbook references) into the methodology/evidence-disclosure file and keep only the universal top-level rules inline.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is a dense operational index with no concept-explainer padding, but it overlaps across the Phase-4 playbook table, the '数据资产规模' count table, and the '引用 / 跨链结构' directory tree, and includes a non-functional 'CLI 助记前缀' section ('当前未实现 CLI, 仅作命名约定'). It is mostly efficient yet could be tightened, matching level 2 rather than the lean level-3 anchor. | 2 / 3 |
Actionability | Provides concrete, executable guidance throughout — named tools per phase (amass/subfinder/httpx/naabu/ffuf/linkfinder/subjack), example dorks ('site:target.com inurl:/admin', 'filetype:env'), test headers ('X-Bug-Bounty: <handle>'), playbook entry hints, and a report skeleton — satisfying the 'concrete, specific guidance' standard for an instruction/router skill. | 3 / 3 |
Workflow Clarity | Clear 5-phase sequence (intake → recon → enum → hunt → report) with per-phase task lists, compliance red-lines as guardrails for destructive ops ('仅 1–3 个 PoC 包, 立即停止'), and a MUST-pass '任务完成自检' checklist plus evidence-discipline reference providing validation checkpoints — clearing the destructive-ops cap. | 3 / 3 |
Progressive Disclosure | SKILL.md is an overview/router pointing one level deep to verified-real references (19 playbooks, 6 methodology, industry, dictionaries, templates, tools) via clearly signaled markdown links and a directory tree; content is appropriately split with easy navigation, matching the level-3 anchor. | 3 / 3 |
Total | 11 / 12 Passed |