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.
A well-structured, highly actionable reference: executable gws commands, validation checkpoints before risky operations (doctor, --validate, --dry-run, schema inspection), and honest handling of a pre-v1.0 dynamically-generated CLI surface. The main weaknesses are repeated verify-caveats that could be consolidated, and a monolithic single-file layout whose referenced scripts/*.py files are absent from the bundle.
Suggestions
Ship the five referenced scripts (gws_doctor.py, auth_setup_guide.py, gws_recipe_runner.py, workspace_audit.py, output_analyzer.py) in a scripts/ directory — every "python3 scripts/..." invocation in the body currently points to a file absent from the bundle, breaking the Quick Start, authentication, and audit workflows.
Move the tabular detail — the environment-variable table, the per-service Required Scopes section, and the Audit Checks catalog — into one-level-deep reference files (e.g. references/auth.md, references/scopes.md, references/audit-checks.md) linked from the relevant sections, keeping SKILL.md as a lean overview.
Consolidate the ~15 inline "(verify)" / "check exact flags with ... --help" caveats into the single top callout, and either complete the placeholder helper examples ("gws sheets +read ...") with representative flags or point them at one commands reference file.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with commands and tables and assumes Claude's competence — no padding explaining what Gmail or OAuth is. The one repeated element is the inline "(verify)" / "check exact flags with ... --help" caveat, which appears roughly 15 times and could be consolidated into the single callout at the top. Efficient with minor trimming opportunities, matching anchor 4 rather than 5. | 4 / 5 |
Actionability | Most commands are copy-paste executable ("npm install -g @googleworkspace/cli", "gws gmail +send --to 'team@company.com' --subject 'Weekly Update'", the doctor/audit/analyzer invocations). Some examples are placeholders ("gws gmail +reply ...", "gws sheets +read ...", '"timeMin": "...""), but the flexibility is explicitly justified by the pre-v1.0 "Verify before scripting" callout, so this stays at anchor 4 with minor gaps rather than dropping to 3. | 4 / 5 |
Workflow Clarity | Workflows are clearly sequenced with validation checkpoints: install is verified via "python3 scripts/gws_doctor.py", auth is a numbered Step 1/2/3 with "--validate", bulk operations mandate "Use --dry-run first", and schema inspection ("gws schema drive.permissions.create") precedes destructive changes — so the destructive/batch cap is not triggered. Minor gaps keep it at anchor 4: the Gmail bulk example ends with "Then feed ids to gmail users messages modify (see ...)" without completing the loop. | 4 / 5 |
Progressive Disclosure | Section structure is solid (Quick Start, Installation, Authentication, four numbered Workflows, Python Tools, Best Practices, Limitations, with a script index table), but everything is inlined in one ~370-line file — the env-var table, per-service scope tables, and audit-check catalog are natural candidates for one-level-deep reference files, and none exist. Critically, the bundle contains no scripts/, references/, or assets/ directories, so the five "python3 scripts/*.py" invocations referenced throughout point to files not present in the bundle. This matches anchor 3 (content that should be separate is inline) rather than anchor 4 (most content appropriately placed). | 3 / 5 |
Total | 15 / 20 Passed |