Content
71%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a lean, well-structured instruction-only skill with concrete env vars, endpoints, and MCP CLI commands, plus explicit validation and confirmation checkpoints. Its main weakness is that the core API flows (token acquisition, webhook sends) are named but not shown with executable examples or error-recovery steps.
Suggestions
Add a short executable example for the most common flow, e.g., a curl snippet obtaining a tenant_access_token from https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal using the env vars, and a minimal signed webhook POST.
Add an error-recovery feedback loop after the verification step (e.g., 'If the dry run or read-back fails, fix the payload and re-verify before sending').
Remove or fold the 'Common Use Cases' section into the Workflow/description to trim redundant tokens.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean with no concept over-explanations — it jumps straight to endpoints ('open.feishu.cn' vs 'open.larksuite.com'), env var names, and CLI commands, assuming Claude's competence. It is not 5 because the 'Common Use Cases' bullet list largely duplicates what the description and Workflow sections already convey and could be trimmed. | 4 / 5 |
Actionability | Concrete, executable specifics are present: exact env vars (FEISHU_APP_ID, FEISHU_APP_SECRET, FEISHU_WEBHOOK_URL), exact CLI commands ('codewhale mcp add', 'codewhale mcp validate', 'codewhale mcp tools'), and named MCP tools. It is not 5 because the core API operations (obtaining a tenant access token, sending a signed webhook message) have no example request or payload, leaving minor gaps in executable coverage. | 4 / 5 |
Workflow Clarity | The 6-step workflow is clearly sequenced and includes explicit checkpoints: 'Verify with a dry run, sandbox document, or read-back call before sending externally visible messages', 'codewhale mcp validate', and the confirmation gate before external writes. It is not 5 because there is no error-recovery feedback loop (what to do when validation fails), and the verification step itself is a direction rather than a concrete check. | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines with well-organized sections (Ground Rules, Common Use Cases, Workflow), no inlined bulk that belongs in separate files, and no external references required. Per the rubric's simple-skill guideline, this earns a 5 with just well-organized sections. | 5 / 5 |
Total | 17 / 20 Passed |