AB promoter-test
Generate an evals file for a drafted skill and measure whether its trigger description fires on the right requests: five to eight phrases that should trigger it, five that should not, three golden input and output pairs, then precision and recall against the description, with a rewrite loop when precision is below 0.8. Use when asked to test this skill's triggers, write evals for the new skill, check the description fires correctly, or as the third step of the promote loop. Produces an evals.json beside the skill, a case in the library's evals, and a precision and recall report.
Generate an evals file for a drafted skill and measure whether its trigger description fires on the right requests: five to eight phrases that should trigger…
As a process B 69/100 · Nearly there — weak spots: when it triggers, failures and branches, running it twice
How to improve
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 2. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
✓ No remarks against the Agent Skills spec
Process rating: all ten parameters 69/100
- 0Failures and branches. Linear process with no failure handling
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 2 mutating operations with no state check
- 60Result and completion. Output format stated, no completion criterion
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 25 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 997 tokens
- 100Progress reporting. Reports progress
- low The response is described with custom markup (3 tags): a typed call is more reliable
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +1No license
- +2Single-language instructions
- +3Description length 585: enough signal without eating the budget
- +4Structure: 10 headings
- +3Step-by-step instructions: 25 items
- +3Output format is stated explicitly
- +4Has examples (1 code blocks)
- +3All 1 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 90.