AA conversation-distill
At the natural end of a meaningful conversation, show a one-line soft reminder asking the user if they want to distill — do NOT auto-start. Only ask when the conversation had distillation value (decisions, insights, judgments, lessons, open questions, action items). If user says yes, run the full 5-step classify→confirm→write flow. Trigger reminder when: (1) closing phrases detected ('thanks', 'done', 'that's all', '好的就这样') AND conversation had substantive content; (2) user explicitly says 'distill', 'wrap up', '收尾'. Never auto-start without asking first.
As a process A 85/100 · Runs to the end — weak spots: 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: 4. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
frontmatter-keyunknown frontmatter key "permissions"
Process rating: all ten parameters 85/100
- 30Running it twice. 5 mutating operations with no state check
- 60Result and completion. Output format stated, no completion criterion
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 45 steps
- 100Failures and branches. 4 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 2820 tokens
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
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)
- -5TODO / placeholder text left in the skill
- -224 emoji in the instructions: noise for the model
- +2Single-language instructions
- +3Description length 561: enough signal without eating the budget
- +4Structure: 20 headings
- +3Step-by-step instructions: 45 items
- +3Output format is stated explicitly
- +4Has examples (3 code blocks)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 80.