orchestration/grill-me
Interview the user relentlessly about a plan or design, resolving every branch of the decision tree before producing supporting docs. Specialised for the Tessl monorepo stack. Use when user says "grill me", "interview me about this plan", "help me think through this design", or similar.
68
86%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
89%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong description with excellent trigger terms and clear 'Use when' guidance. The interactive interviewing concept is distinctive and well-articulated. The main weakness is that the specific capabilities could be more concrete—what types of supporting docs are produced, and what does the decision tree resolution process look like in practice.
Suggestions
Specify the types of supporting docs produced (e.g., 'producing design docs, ADRs, or implementation plans') to improve specificity.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description names the domain (plan/design interviewing) and mentions producing 'supporting docs', but doesn't list specific concrete actions beyond interviewing and document production. What kind of docs? What does 'resolving every branch of the decision tree' concretely entail? | 2 / 3 |
Completeness | Clearly answers both 'what' (interview the user about a plan/design, resolve decision branches, produce supporting docs) and 'when' (explicit 'Use when' clause with specific trigger phrases). | 3 / 3 |
Trigger Term Quality | Includes very natural trigger phrases users would actually say: 'grill me', 'interview me about this plan', 'help me think through this design'. These are realistic, varied, and cover the main ways a user would invoke this skill. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with unique trigger terms like 'grill me' and the specific interactive interviewing approach. Specialization to the Tessl monorepo stack further narrows the niche. Unlikely to conflict with other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, highly actionable skill that provides a thorough interview framework with clear phasing, validation checkpoints, and concrete codebase-specific guidance. Its main weakness is length — the document packs a lot of detail inline that could benefit from progressive disclosure into supporting files, and some framing sentences could be trimmed. The workflow design is excellent, with explicit gates preventing premature completion.
Suggestions
Extract codebase-specific patterns (TanStack Router loader pattern, openapi-react-query return contract, route registration, codegen steps) into a separate `CODEBASE_PATTERNS.md` reference file to reduce the main skill's token footprint.
Trim explanatory framing sentences like 'These documents are the single source of truth for ticket creation — they must contain everything an autonomous agent needs...' which restate the purpose rather than adding actionable guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is well-structured and mostly efficient, but some sections are verbose with explanatory context that Claude doesn't need (e.g., explaining why specs must be complete, restating principles that are implicit in good interviewing). The parenthetical codebase-specific notes are valuable but could be more tightly formatted. | 2 / 3 |
Actionability | The skill provides extremely concrete, actionable guidance: specific questions to ask in each phase, exact file paths and commands for the codebase (e.g., `bun run api:sync`, route registration in `apps/backend/src/routes/v1/index.ts`), precise output file structure, and a coverage mapping table format. Every phase has clear deliverables. | 3 / 3 |
Workflow Clarity | The five-phase workflow is clearly sequenced with explicit ordering (product intent before architecture, operability before completeness check). Phase 4 serves as a validation checkpoint, Phase 5 is a coverage gap analysis that explicitly states 'do not end the interview until every row has an owner' — a clear feedback loop preventing premature completion. | 3 / 3 |
Progressive Disclosure | The content is well-organized with clear headers and sub-sections, but it's a long monolithic document (~200 lines) with no references to external files. The codebase-specific patterns (TanStack Router, openapi-react-query, route registration) could be split into a separate reference file to keep the main skill leaner. No bundle files are provided to support this. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
Reviewed
Table of Contents