Covers the refusal paths a handler already implements but nothing tests - a batch endpoint that must apply all rows or none, optimistic-concurrency version conflicts between two editors, or a delete that deliberately separates who you are from what you may do from the state the record is in. For each happy-path test, produces companions exercising input validation rejection, missing required fields, type mismatches, authorization failures, rate-limit errors, and adversarial payloads from the malicious-payload-bank, emitted as parameterized tests in the project's runner format. Use when code has deliberate error paths and the suite only proves the success case.
69
87%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Passed
No findings from the security scan
Tessl evals compare success rates of agents with and without our optimized context