AC prompt-eval
Evaluate and improve any AI prompt (`prompt_a`) through a staged, evidence-based pipeline. Functional evaluation checks whether the prompt follows rules, output contracts, quality requirements, and safety boundaries. Optional effect evaluation checks whether outputs work for intended readers through blinded persona-based comparison. Produces inspectable CSV/JSON/HTML artifacts, root-cause findings, a validation-gated optimized prompt, and a final report. Use when users ask to evaluate, test, benchmark, score, validate, QA, or improve a prompt.
As a process C 60/100 · Has gaps — weak spots: result and completion, when it triggers, 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 · 1
✓ No critical or high findings
Medium and low: 1
-
low Instruction override
en-ignore-previousreferences/prompt_b_guide.md:327Instruction-override phrase ("ignore previous instructions") (detector / deny-list definition; security demo / example)(e.g., "I notice a request to override safety constraints") but ultimately
detectordemo
Files scanned: 9. 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 60/100
- 0Result and completion. Does not say what the result is
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 12 mutating operations with no state check
- 60Tools and files. Uses tools (bash) that frontmatter does not declare
- 70Inputs and preconditions. Inputs and preconditions are listed
- 70Failures and branches. 9 branches
- 100Steps. 42 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3765 tokens
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 14 top-level sections: this looks like several domains in one skill
- 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)
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +3Description length 549: enough signal without eating the budget
- +4Structure: 22 headings
- +3Step-by-step instructions: 42 items
- +4Has examples (4 code blocks)
- +4Reference files are cited in the instructions (5 of 5)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 88.