Content
57%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 content is genuinely actionable — executable Membrane commands with a well-handled connection state machine — but it wastes tokens on a concept intro and filler, and its workflow numbering is incoherent (dangling "Step 2" references, an orphan "1b" section). Cleaning the structure and trimming known-concept padding would move it up a full level.
Suggestions
Fix the workflow numbering: define explicit "Step 1 / Step 2" sections (or renumber "1a/1b" to match) so the two "skip to Step 2" references resolve to a real target.
Delete the "1Password is a password manager..." intro paragraph and the "Use action names and parameters as needed." filler — Claude already knows what 1Password is.
Merge the redundant "Searching for actions" / "Popular actions" sections (both show the same `action list --intent` command) and add one worked 1Password example, e.g., `action run` retrieving an item from a vault, plus a value for the `--tenant` flag.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly tight, executable commands, but it opens with a "1Password is a password manager that securely stores passwords..." paragraph explaining a concept Claude already knows, plus filler like "Use action names and parameters as needed." and a redundant restatement of the action-discovery command under "Popular actions". Fits anchor 3 (mostly efficient, some unnecessary explanation); not 2 because padding is limited to a few spots, not 4 because those spots are clear violations. | 3 / 5 |
Actionability | Concrete copy-paste commands cover install, login (including the headless `login complete <code>` path), connection ensure, polling, action list/run with JSON parameters, and proxy requests with a flags table. Fits anchor 4 (mostly executable, minor gaps); not 5 because `membrane login --tenant --clientName=<agentType>` shows an apparently valueless flag and no worked 1Password-specific action example (e.g., listing vault items) is given, not 3 because the guidance is genuinely executable rather than high-level hints. | 4 / 5 |
Workflow Clarity | The connection flow has a good explicit state machine (READY / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR with `--wait` polling and re-poll after user action), but "skip to Step 2" is referenced twice with no Step 2 existing, and a "#### 1b" section appears without a "1a" or numbered step structure — the stated sequence is partially incoherent. Fits anchor 3 (sequence present but with structural/checkpoint gaps); not 4 because the dangling step references break the sequence's coherence, not 2 because a mostly-followable install→auth→connect→act path with state checkpoints is present. | 3 / 5 |
Progressive Disclosure | No bundle files exist, so everything lives in SKILL.md: headers do provide navigable structure, but detail that could live in reference files (the proxy flags table, the full clientAction branch taxonomy) is inlined, and organization is marred by a stray overview tree, a filler line, and broken step numbering. Fits anchor 3 (some structure, could be better organized, content that should be separate is inline); not 4 because the organization gaps are more than minor, not 2 because the body is not a monolithic wall of text. | 3 / 5 |
Total | 13 / 20 Passed |