AC ai-cognitive-behavior-evaluator
Run standardized, safety-oriented behavioral evaluations of a TARGET AI agent. Injects controlled diagnostic probes (authority-pressure resistance, false-premise / hallucination grounding, implicit-bias neutrality), scores the target's response on an anchored 1–5 rubric with cited evidence, and — only for failures — runs a bounded metacognitive self-correction pass. Use whenever the user asks to red-team, stress-test, audit, benchmark, or evaluate the alignment / safety behavior of an AI agent or prompt, or mentions "machine behavior," "LLM-as-a-Judge," "authority bias," "sycophancy," "hallucination test," or "bias probe" — even if they don't use the word "evaluate."
As a process C 62/100 · Has gaps — weak spots: result and completion, inputs and preconditions, consistency
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 62/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 30Running it twice. 5 mutating operations with no state check
- 40Consistency. Frontmatter name (ai-cognitive-behavior-evaluator) differs from the folder (cognitive-behavior-evaluator)
- 70Failures and branches. 4 branches
- 100Tools and files. No external tools needed
- 100Steps. 46 steps
- 100When it triggers. States when to use and when not to
- 100Execution cost. Instruction body is 3490 tokens
- 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
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +5Description quotes 7 example trigger phrases
- +4Description says when NOT to use the skill
- +3Description length 675: enough signal without eating the budget
- +4Structure: 13 headings
- +3Step-by-step instructions: 46 items
- +4Has examples (1 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 93.