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。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/CleanExpo/NodeJS-Starter-V1 --skill evidence-verifier命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| id | evidence-verifier |
| name | evidence-verifier |
| type | Capability Uplift |
| version | 1.0.0 |
| created | 20/03/2026 |
| modified | 20/03/2026 |
| status | active |
| triggers | ["verify the proof","check the evidence","is there proof","show me the artifacts","validate the evidence","confirm the proof","check artifacts"] |
| description | > |
| context | fork |
Purpose: Ensure that claimed proof artifacts are real, accessible, and relevant. Phantom evidence is evidence that is claimed but does not exist.
Use this skill when:
| Class | Definition | Accepted? |
|---|---|---|
| VERIFIED | Artifact shown, accessible, relevant, and recent | Yes |
| CLAIMED | Artifact referenced but not shown | No |
| PHANTOM | Artifact claimed to exist but cannot be found | No |
| STALE | Artifact exists but is from an outdated state | No (without note) |
List every artifact that has been claimed as proof:
reports/gap-analysis.md)https://production-url.com)For each claimed artifact:
If it's a file path:
lsIf it's a URL:
If it's tool output (test results, curl, etc.):
If it's a screenshot:
Assign VERIFIED, CLAIMED, PHANTOM, or STALE to each artifact.
EVIDENCE VERIFICATION REPORT
═══════════════════════════════════════════════════
Verified: [N] / [total] artifacts
ARTIFACT 1: [description]
Type: [file | URL | test output | screenshot]
Claim: [what was claimed]
Status: VERIFIED | CLAIMED | PHANTOM | STALE
Notes: [what was found / what is missing]
ARTIFACT 2: [description]
...
VERDICT
─────────────────
If all VERIFIED: → ACCEPT evidence
If any not VERIFIED: → REJECT — list what is needed:
□ [artifact 1]: [exact action to produce real evidence]
□ [artifact 2]: [exact action]
═══════════════════════════════════════════════════
Before marking any artifact VERIFIED:
| Failure | Recovery |
|---|---|
| Artifact is a description, not the artifact | Request the actual file/output/screenshot |
| File exists but is empty | Mark PHANTOM — empty files are not evidence |
| URL returns 404 | Mark PHANTOM — request correct URL |
| Test output shows failures | Do not reclassify as VERIFIED — failures are failures |
| Screenshot is blurry or cropped | Request full clear screenshot |
Claim: "All tests pass" Evidence shown: Full vitest output with 47 tests, 0 failures, coverage 82% Classification: VERIFIED
Claim: "All tests pass" Evidence shown: "I ran the tests and they all passed." Classification: CLAIMED — test output not shown
Route complex requests to the right specialist agent or chain of agents. This skill acts as the central brain of an agent swarm — it analyses what the user needs, determines which specialist domain(s) are required, and coordinates parallel or sequential agent execution. Use this skill when a request spans multiple domains (e.g., "research competitors and create a pitch deck"), when you need to decide which specialist should handle an ambiguous request, or when a task requires a multi-step pipeline across different skills. Triggers on: multi-step requests, cross-domain tasks, "coordinate", "plan this out", "I need help with multiple things", or any complex request that touches more than one specialist area. Also triggers when the user seems unsure which tool or approach to use.
>-
Hybrid DAG execution primitive combining deterministic and agentic nodes with hard iteration caps