Content
61%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 strongly actionable CLI guide with copy-paste commands and a genuinely well-handled connection-state loop, including error-recovery guidance. Its main weaknesses are the boilerplate intro, a broken 'Step 2' cross-reference, and the absence of any validation guidance for the destructive and batch operations it advertises.
Suggestions
Delete the 'Bubble is a no-code development platform...' intro paragraph and the sentence fragment 'When to use which actions: Use action names and parameters as needed.' — Claude already knows what Bubble is.
Fix the internal navigation: either add explicit numbered 'Step 1 / Step 2' headings or remove the 'skip to Step 2' and '1b' references that point to headings that don't exist.
Add a validation checkpoint for destructive/batch operations, e.g. verify a record exists with get-record before delete-record, or confirm the count before bulk-create-records (max 1000), and check the response's error field before proceeding.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The opening paragraph explains what Bubble is ("Bubble is a no-code development platform that allows users to build web applications without writing any code. It's used by entrepreneurs, startups...") — a concept Claude already knows — and fragments like "When to use which actions: Use action names and parameters as needed." and the awkwardly spaced "**Agent Types** : claude, openclaw, codex..." padding could be cut. Matches anchor 3 (mostly efficient, some unnecessary explanation); not 4 because multiple sections need tightening. | 3 / 5 |
Actionability | Commands are concrete and copy-paste ready throughout (install, `membrane login`, `membrane connection ensure`, polling with `--wait`, `action list/run`, proxy flag table). Matches anchor 4 (mostly executable, minor gaps) rather than 5 because `action run` shows only placeholders (`<actionId>`, `'{"key": "value"}'`) and `membrane request` lacks a worked example with a real Bubble endpoint. | 4 / 5 |
Workflow Clarity | The connect-poll loop has real checkpoints (READY / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR with follow-up polling), but the body says "skip to **Step 2**" while no Step 2 heading exists, and the destructive/batch actions it documents (delete-record, replace-record, bulk-create-records up to 1000 records) have no validation or verification steps — which caps this dimension at 3. | 3 / 5 |
Progressive Disclosure | No bundle files exist, and the ~150-line body is cleanly sectioned with headers, an action table, and a proxy flag table, with no nested references and external links clearly placed ("Official docs"). Minor gaps (the orphaned '1b'/'Step 2' numbering, slight redundancy between 'Searching for actions' and 'Best practices') keep it at anchor 4 rather than 5. | 4 / 5 |
Total | 14 / 20 Passed |